Introduction
Copyright law is a law of enforcement as much as of entitlement. A right that cannot be practically vindicated confers little more than a formal title, and nowhere is the distance between title and remedy wider than in the digital environment, where a work can be copied, uploaded and monetized by thousands of dispersed users within hours.1 Tanzania’s creative economy has expanded rapidly on the back of mobile internet use, streaming services and social media, giving musicians, filmmakers, photographers and writers access to audiences that were unreachable without traditional intermediaries.2 The same technologies have exposed a structural weakness in the statutory framework that is meant to protect them.
The Copyright and Neighbouring Rights Act, Cap. 218 R.E. 2023 (the Act), grants authors exclusive rights of reproduction, distribution, communication to the public and making available,3 and criminalizes their unauthorized exercise.4 These rights are framed in language broad enough to capture uploading, sharing and streaming. What the Act itself does not contain is any provision on the liability of the platforms through which online infringement is committed, any procedure by which a rights holder may secure the removal of infringing content without commencing full civil proceedings, or any copyright-specific mechanism for cross-border notice-and-takedown or intermediary compliance.5 It would nevertheless be wrong to conclude that Tanzanian law offers no route against intermediaries. Part V of the Cybercrimes Act, 2015 gives service providers conditional immunities (sections 39 to 44), allows any person to notify a service provider of infringing material (section 45) and makes the use of a computer system to violate intellectual property rights an offence (section 24).6 The Cybercrimes (General) Regulations, 2016 prescribe a take-down notification form that lists ‘infringing copyrights’ and require a service provider that is satisfied with a notification to act on it within twenty-four hours (regulation 7).7 The Online Content Regulations, 2020 add a duty on online content hosts to ‘ensure that prohibited contents are removed upon notification by the Authority or affected party’ (regulation 15), which a rights holder can use as an affected party because the Schedule lists content infringing intellectual property rights as prohibited, and duties on licensees to ensure compliance with applicable laws, to monitor and moderate user generated content to avoid publication of prohibited content and to remove it when ordered (regulation 9, as amended in 2022), duties that the 2025 amendments extend to social media platform owners (regulation 15A).8 These regulatory duties concern online content generally and do not themselves establish a copyright-specific notice-and-counter-notice process. The question is therefore not whether Tanzania has a notice-and-takedown route but whether the route it has is fit for copyright. The United States has conditioned platform immunity on notice-and-takedown since 1998,9 and within Tanzania’s regional neighborhood Kenya,10 Nigeria11 and, most recently, Uganda12 have each legislated for intermediary obligations in their copyright statutes, Kenya and Nigeria with a counter-notice and civil liability for a provider that does not comply.
In the two cases examined, courts found infringement, but in one the appeal was still open after long delay, and in both against defendants who could be identified and served. In Multichoice Tanzania Limited v Maimuna K Kiganza the High Court upheld a finding that a broadcaster had infringed the respondent’s film by airing it on a digital satellite channel without authorization.13 In the dispute between the artistes known as MwanaFA and AY and a telecommunications operator over the alleged unauthorized commercial use of their songs as caller tunes, proceedings begun in 2012 reached the Court of Appeal in 2025 on a threshold question of jurisdiction.14 If a dispute of that kind, in which the trial court had found the use unauthorized and the Court of Appeal in 2025 sent the remaining grounds of appeal back to the High Court, can take thirteen years to settle a preliminary question, it is difficult to describe the framework as equipped for infringement that can cause irreversible commercial harm within days.
This article therefore asks three questions. First, is the legal framework, comprising the Act together with the cybercrime and online-content instruments that apply to intermediaries, adequate to regulate copyright protection against platform-mediated online infringement? Secondly, are the institutional and enforcement mechanisms available to rights holders, principally the Copyright Office of Tanzania (COSOTA), effective for that purpose? Thirdly, what do the experiences of Kenya, Nigeria, South Africa, Uganda and the principal global benchmarks teach about the design of a reform that Tanzania could realistically adopt? The study adopts a doctrinal legal research methodology, and each jurisdiction is examined along five parameters, which correspond to the elements of a copyright-appropriate intermediary regime against which the Tanzanian framework is assessed: (i) whether a defined category of intermediary is recognized and a conditional immunity conferred; (ii) whether a notice-and-takedown procedure with a fixed timetable exists; (iii) whether a counter-notice or appeal safeguard protects users; (iv) the institutional locus of administration and adjudication; and (v) the availability of copyright-specific cross-border notice-and-takedown or compliance arrangements.
The argument advanced is that the framework provides a general legal route for takedown and conditional intermediary protections, but that its statutory design lacks a coherent copyright-specific procedure with clear standards, counter-notice safeguards and an expressly assigned institutional role. The claim is one about statutory design, established by doctrinal analysis; it is not a claim about how institutions perform in practice, for which this study collected no data. Tanzania has a notice-and-takedown route and a scheme of conditional immunities, but they sit in cybercrime and online-content instruments that were not designed for copyright, the consequences they attach to a provider’s inaction are criminal rather than civil, and they omit what a copyright-appropriate regime requires: a counter-notice procedure, objective standards for removal, a defined role for the Copyright Office (in Tanzania, COSOTA) and a civil compliance and liability consequence for a provider that ignores a copyright notice. The Act itself already provides civil remedies against infringers, and the argument concerns the absence of a civil consequence for the intermediary, not the absence of civil remedies as such. This is a narrower and more consequential claim than a general finding of inadequacy, and a different claim from the one that Tanzanian law is silent. It matters because it dictates the order of reform: a capacity-centered diagnosis points towards training and budgets, whereas a mechanism-centered diagnosis points to the design and placement of the legal mechanisms as the necessary first step, on which capacity-building can then operate.
The article proceeds as follows. Part II sets out the background to the problem. Part III develops the theoretical framework and Part IV reviews the literature. Part V presents the international, domestic and comparative legal frameworks, Part VI sets out the findings and Part VII discusses them. Part VIII concludes and makes recommendations.
Background to the problem
Copyright has always developed in response to the technology of reproduction. The earliest regimes emerged with the printing press and were concerned with the control of physical copying;15 later extensions to photography, sound recording and broadcasting preserved a central focus on tangible copies and identifiable modes of distribution.16 Enforcement mechanisms were correspondingly designed around territorially confined acts and identifiable infringers. Digital technology removed each of those assumptions by enabling instantaneous, low-cost and near-perfect reproduction, and global dissemination without centralized intermediaries.17
Tanzania’s Act reflects the earlier model. Its core provisions are medium-neutral and, in principle, capable of extending to digital uses, but its enforcement architecture presupposes a single infringer against whom a civil or criminal remedy can be sought.18 The Act itself has no counterpart to the notice-and-takedown mechanism that, since the Digital Millennium Copyright Act 1998, has given United States rights holders a remedy that runs against the entity actually capable of removing content within hours, without first requiring them to identify and sue the uploader.19 A counterpart of a different kind sits elsewhere in the statute book, in the Cybercrimes Act and the regulations made under it, but it was drafted for unlawful content generally and not for copyright (Part V). The result is a regulatory lag, and a regulatory misfit, in which the pace of technological change has outstripped the evolution of copyright-specific legal control.
The international position adds a distinctive complication. The WIPO Copyright Treaty 1996 introduced the digital-specific rights of communication to the public and making available,20 on which the Act’s definition of communication to the public is evidently modeled. Yet Tanzania is not a contracting party to that Treaty.21 Its statute therefore borrows the vocabulary of an instrument it has never joined, with the consequence that it enjoys none of the interpretive guidance, peer accountability or monitoring that membership would bring, while still bearing the practical consequences of operating a framework styled on that instrument’s assumptions.
Meanwhile the region has moved. Kenya amended its Copyright Act in 2019 to insert an express regime governing internet service providers;22 Nigeria’s Copyright Act 2022 devotes a part of the statute to online content;23 and Uganda enacted the Copyright and Neighbouring Rights (Amendment) Act 2026, which inserts a section 49A empowering the Registrar of Copyright or a copyright owner to issue a notice requesting a service provider to block or take down infringing content.24 Tanzania is not an outlier in having legislated on the liability of service providers, since it did so in 2015, before Kenya and Uganda. It is an outlier in having no copyright-specific regime and in relying on general instruments that lack the safeguards of the regional models. That divergence has direct implications for the protection of creators and for the credibility of the East African Community’s commitment to harmonized intellectual property standards.
Theoretical framework
Copyright scholarship offers no single justification for the institution. Fisher observes that modern systems draw eclectically on natural-rights, personality and utilitarian arguments,25 and the Act reflects that eclecticism by protecting both economic rights of a utilitarian character and moral rights grounded in personality. For the purposes of an enforcement inquiry, three theoretical perspectives are particularly productive, and together they supply the evaluative standard applied in this article.
The enforceability thesis. The utilitarian or incentive account, associated with Landes and Posner, justifies copyright as an instrument for correcting the under-production of works that would otherwise be costlessly copied once disclosed.26 The theory is concerned with calibrating protection so as to balance the incentive to create against the social cost of restricted access.27 Its significance for the present inquiry is that the incentive function depends on the practical enforceability of exclusive rights, not merely their formal existence. An author who cannot secure the removal of infringing content from a dominant platform experiences the same disincentive as one who holds no right at all. Where the capacity to enforce against platform-based infringement is weak, as in Tanzania, the incentive rationale for copyright is correspondingly diminished even though the entitlement remains intact.
The least-cost-avoider rationale. Calabresi’s principle holds that liability should presumptively fall on the party able to prevent a harm at the lowest cost.28 Landes and Lichtman apply the logic to online infringement: individual rights holders face prohibitive costs in monitoring millions of dispersed uploads, whereas the hosting platform is placed to detect or remove infringing material at scale.29 This explains why intermediary regimes direct obligations at platforms rather than leaving liability to accumulate against anonymous or judgment-proof direct infringers, and it explains why liability is conditioned on knowledge and notice rather than imposed strictly. It also anticipates the doctrinal lineage of the safe harbor, which developed from the distinction between a passive conduit and an actor exercising control over infringement.30
The democratic-culture constraint. Netanel’s social-planning account treats copyright as an implicit subsidy for an independent sector of creative production on which democratic civil society depends.31 On this view the objective of copyright policy is not only to maximize the private return to authors but to sustain a diverse public sphere.32 This perspective supplies the check on how enforcement should be designed. A takedown regime calibrated only to the interests of rights holders, without counter-notice and due-process safeguards, risks the over-removal and chilling effects documented in comparable regimes elsewhere.33 The risk is acute in Tanzania, where social media and video-sharing platforms have become primary sites of political commentary, satire and civic discourse.
Taken together, the three perspectives yield a dual standard. A reform proposal must be assessed not only by whether it strengthens enforcement, which the first two perspectives require, but also by whether it does so without displacing the expressive and civic uses of platforms that the third perspective identifies as independently valuable. This dual standard is applied throughout the findings, discussion and recommendations that follow.
Literature review
The literature on copyright in the digital environment is organized around the tension between analogue doctrine and platform-based dissemination. There is broad agreement that technology has strained existing frameworks, but disagreement on whether existing rules suffice or new mechanisms are required, particularly in developing jurisdictions. The works reviewed below are grouped by the question they address.
Cornish, Llewelyn and Aplin provide a comprehensive doctrinal exposition in which rights of reproduction and communication to the public extend to digital acts including online dissemination and streaming.34 The analysis supports the premise that section 9 of the Act could, in principle, apply to online infringement. The authors do not, however, engage with jurisdictions in which enforcement capacity is limited or intermediary regulation absent, so their doctrinal optimism does not resolve the practical gap experienced in Tanzania.
Bainbridge and Rose trace the adaptation of copyright to successive reproduction technologies, from the printing press to digital media.35 The Tanzanian experience fits that reactive pattern, but the historical account alone is incomplete: the Tanzanian challenge is structural rather than merely temporal, because the law lags not only technologically but institutionally, with limited judicial precedent on online infringement and weak enforcement infrastructure.
Goldstein and Hugenholtz emphasize the role of international frameworks, particularly the WIPO Internet Treaties, in producing convergence around technologically neutral rights.36 Their analysis shows normative convergence but not that convergence translates into operational enforcement capacity. In Tanzania, treaty participation is partial and its domestic implementation largely implicit, which opens a gap between norm adoption and legal effectiveness. Samuelson adds that digital media collapse the traditional distinction between distribution and communication, requiring courts to interpret long-standing statutory categories flexibly,37 a point of direct relevance to how Tanzanian courts might read section 9.
Magessa, the most recent contribution, examines the adequacy of Tanzania’s copyright framework and documents resource constraints, insufficient training and limited technological infrastructure at COSOTA and in the judiciary, citing Multichoice as illustration.38 That work frames the problem predominantly as one of institutional capacity. It does not, however, test whether capacity is the operative constraint by asking what legal powers a properly resourced institution would require.
Urban and Quilter supply the empirical caution that takedown systems carry real risks of over-removal.39 Their study is drawn from the United States and does not address the enforcement constraints typical of developing jurisdictions.
Four gaps emerge. Much doctrinal scholarship assumes functional enforcement systems or does not interrogate the statutory sufficiency of the Act. There is limited integration between doctrinal analysis of the Act and the practical realities of platform enforcement. The comparative works reviewed are drawn from the United States and Europe and do not address the institutional constraints of developing jurisdictions. And the most recent Tanzanian scholarship reviewed has not undertaken a combined doctrinal, institutional and comparative evaluation. This article fills those gaps by testing the capacity-centered account against the statutory text and the comparative record.
Comparative legal framework
A. The international and regional framework
Tanzania is bound by genuine international copyright obligations. As a member of the Berne Union40 and a Member of the World Trade Organization subject to the TRIPS Agreement,41 it must extend national treatment and minimum substantive protection to domestic and foreign rights holders alike. Neither instrument was drafted with platform-mediated exploitation in mind. TRIPS Part III requires enforcement procedures that permit effective action against infringement, including expeditious remedies,42 which bears on the adequacy of the domestic route, but neither instrument prescribes intermediary liability or notice-and-takedown.
The next layer complicates rather than clarifies. The WIPO Copyright Treaty and the WIPO Performances and Phonograms Treaty supply the principal international benchmark for digital-era exclusive rights, including the making-available right.43 Tanzania is a member of WIPO and of the Berne Union but is not a contracting party to the Copyright Treaty, which has been in force since 2002.44 Tanzania’s constitutional posture is dualist: treaties are ratified by Parliament and take domestic effect through legislation where their implementation requires it.45
The regional and continental layers do not close the gap. The African Regional Intellectual Property Organization (ARIPO), of which Tanzania is a member,46 has adopted no protocol on online copyright enforcement or intermediary liability; its Swakopmund Protocol concerns traditional knowledge and folklore, and its Kampala Protocol the voluntary registration of copyright and related rights.47 The East African Community Treaty commits partner states to cooperate on intellectual property in general terms,48 but no binding Community instrument harmonizes copyright protection or online enforcement. The African Continental Free Trade Area must be taken in two parts. Tanzania is a party to the Agreement Establishing the AfCFTA,49 but the Protocol on Intellectual Property Rights, adopted in February 2023, is the instrument that extends to copyright and related rights and encourages effective enforcement against digital infringement,50 and it has not entered into force, because it requires twenty-two ratifications and has not yet attracted them.51 It therefore creates no binding obligation for Tanzania, whatever Tanzania’s own position on ratification of the Protocol.
B. Tanzania’s domestic framework
The Constitution does not contain a dedicated copyright provision, but article 24(1) guarantees the right to own property, a guarantee capable of extending to the rights the Act confers.52 The Act itself defines key terms such as ‘communication to the public’ and ‘reproduction’ in technologically neutral language,53 vests the exclusive rights in section 9(1) and criminalizes their unauthorized exercise in section 45. Read in isolation, the text is not deficient: uploading is a reproduction, sharing is a distribution and streaming is a communication to the public.
The difficulty lies in the enforcement provisions. A close reading of sections 39 to 47 and 51 shows civil remedies of damages, an account of profits and injunctive relief; criminal offences for commercial infringement; border measures empowering customs authorities to detain infringing goods; and a function in COSOTA of dealing with issues of anti-piracy.54 The civil and criminal remedies are medium-neutral but specify no procedure by which they can be invoked against a platform rather than an identified defendant. The border measures are structurally inapplicable to purely digital infringement, since there are no goods for customs authorities to detain. COSOTA’s anti-piracy function is broad enough in principle to support a digital enforcement mandate, and no instrument or published record examined shows it exercised to create one. This article does not examine the general law on jurisdiction, mutual legal assistance and international cooperation. Under the Act before its 2019 amendment, jurisdiction over copyright disputes was vested exclusively in the District Court irrespective of the value of the claim, a rule confirmed by the Court of Appeal in 2025; the Act now refers to the court of competent jurisdiction.55
Three adjacent bodies of law intersect with the problem, and the first two already supply the nucleus of a notice-and-takedown route. The Cybercrimes Act, 2015 is the more important. Part V confers conditional immunities on access, hosting, caching, hyperlink and search-engine providers (sections 40 to 44), and provides that a provider shall not monitor the data it transmits or stores (section 39(1)). A hosting provider is immune for information stored at a user’s request if it removes or disables access immediately on an order of a competent authority or court, or informs the relevant authority when it becomes aware of illegal information (section 41(1)); a provider that has knowledge of illegal information or activity must remove it, suspend the service concerned and notify the appropriate authority (section 39(4)). The rights holder’s own notice acquires force through section 45, under which any person may notify a service provider, by a take-down notification, of data or activity infringing the rights of a third party, of unlawful material or of any matter contrary to written law. The notification must be in a permanent medium and must give the complainant’s name, address and signature, identify the right allegedly infringed and the material complained of, state the remedial action required and contain statements of good faith and accuracy (section 45(2)). A knowingly false notification is an offence (section 45(3)); a provider that fails to act is exposed to being charged as though it had itself originated the content (section 45(4)); a complainant whose notification is ignored may report the failure to a competent authority, which may take the material down or order the provider to act (section 45(5)); and a provider is not liable for a take-down done in compliance (section 45(6)). Section 24 separately makes it an offence to use a computer system with intent to violate intellectual property rights protected under any written law, with a minimum fine or term of imprisonment and, in commercial cases, an order to compensate the victim.56
The Cybercrimes (General) Regulations, 2016 make that route operational. Any person who becomes aware of infringing or unlawful material may file a take-down notification on the prescribed form, which offers, among its grounds, ‘infringing copyrights to which I am entitled’, and must furnish it to the service provider (regulation 6). A provider that is satisfied with the notification must take the material down within twenty-four hours of receipt and, if it has reasonable grounds not to do so, must give its reasons to the complainant or the competent authority within the same period (regulation 7).57 The Online Content Regulations, 2020, made under the Electronic and Postal Communications Act, add a second and separate layer. The Tanzania Communications Regulatory Authority (TCRA) may direct the removal of prohibited online content; online content hosts must adopt a code of conduct for hosting contents and ‘ensure that prohibited contents are removed upon notification by the Authority or affected party’ (regulation 15), and since paragraph 6 of the Schedule of prohibited content lists content that infringes intellectual property rights, a rights holder can use regulation 15 as an affected party; licensees must ensure that online content complies with applicable laws, ‘monitor and moderate user generated content to avoid publication of prohibited content’ and ensure that prohibited content is removed immediately when ordered by the Authority (regulation 9, as amended by GN No. 136 of 2022); and the 2025 amendments extend a corresponding duty to social media platform owners, who must deploy a mechanism for filtering and removing prohibited content (regulation 15A).58 A further amendment, the Online Content (Amendment) Regulations, 2026, in operation from 1 July 2026, is concerned, on the text and commentary examined, with licensing categories and fees for online media services, and does not appear to alter regulations 9, 15 or 15A.59
These instruments do not amount to a copyright regime, for five reasons. First, they are general: a notification may allege any unlawful matter and regulation 15 covers all prohibited content, so nothing in them is calibrated to the exceptions and limitations that qualify exclusive rights. Secondly, they supply no counter-notice: neither section 45, regulation 7 nor regulation 15 gives the user whose material is removed a right to be told, to respond or to have the material restored, and regulation 7(2) addresses only the provider’s reasons for declining to act. Thirdly, they supply no objective standard: regulation 7 leaves removal to the provider being ‘satisfied’ with the notification, which itself asks only for statements of good faith and accuracy, and regulation 15 prescribes no form, period or standard. Fourthly, they give COSOTA no defined role: the Cybercrimes Act speaks of a ‘competent authority’, which the Regulations define as any person or organization with jurisdiction and due legal authority over the matter, and no instrument examined here says whether COSOTA is one, so a rights holder cannot know whether a provider’s failure to act may be reported to it. Fifthly, the consequences for a provider’s inaction are criminal in character: a provider that fails to act risks being charged, a false notifier commits an offence, and the uploader faces the minimum penalties of section 24. The two layers also pull in different directions. Section 39(1) of the Cybercrimes Act provides that a provider shall not monitor the data it transmits or stores, whereas regulations 9 and 15A of the Online Content Regulations require monitoring and filtering, and the removal duties of the Online Content Regulations run on a separate footing from the twenty-four hours in regulation 7. The National Arts Council adds a further, only partially overlapping, layer of regulatory authority over the performing arts.60 A rights holder confronting online infringement must therefore work across three uncoordinated regimes, administered by different institutions, none of which was designed for the purpose.
C. The judicial record
Two decisions test the framework in practice. In Multichoice, the High Court rejected the appellant’s argument that it merely provided subscriber-management services and held it liable for airing the respondent’s film without authorization.61 The decision is encouraging, since it shows that the courts will read the exclusive rights generously enough to reach sophisticated digital exploitation, and it is conceptually significant because it applies, in a broadcasting context, the distinction between passive transmission and active exploitation that underlies intermediary-liability doctrine elsewhere. But the respondent’s only route to a remedy was full civil litigation against a single, identifiable defendant. Nothing in the case suggests that an equivalent remedy would be available against an anonymous uploader or an offshore platform, where identification and service are themselves the principal obstacles.
The caller-tune litigation, Mwinyijuma, is more troubling. The suit was filed in 2012; the District Court decided the matter in 2016, finding the use unauthorized, and the Court of Appeal in 2025 confirmed that, under the Act before its amendment, the District Court had exclusive jurisdiction over copyright matters regardless of the value of the claim, and sent the remaining grounds of appeal back to the High Court.62 Thirteen years to resolve a procedural preliminary, in a case where the trial court had found the use unauthorized, is not a delay that any framework adequate to digital-era enforcement could tolerate.
D. The comparator frameworks
Kenya. The Copyright (Amendment) Act 2019 inserted sections 35A to 35D into the Copyright Act 2001. They create a conditional safe harbor for providers performing conduit, caching, hosting and information-location functions, available where the provider does not select or modify the material, does not promote the infringing content and, on receipt of a compliant takedown notice, removes or disables access within forty-eight business hours.63 The provider must notify the person responsible for the content and give them a copy of the notice; it is to disable access unless it receives a counter-notice contesting the notice; and a provider that fails to act without valid justification is liable for any resulting loss or damages (section 35B(4) to (6)).64 A notice must identify the complainant and the work and be supported by a good-faith declaration; a malicious notice is an offence, and a court may order disclosure of an infringer’s identity.65 Kenya’s regime is not presented as an unqualified model. A 2021 amendment bill proposed repealing the takedown and service-provider provisions,66 but sections 35A to 35D remain in the Act as revised to 2022,67 and commentary identifies continuing uncertainty over the standard a provider is to apply in adjudicating between a notice and a counter-notice.68
Nigeria. The Copyright Act 2022 devotes Part VII to online content, covering notice and takedown, a procedure for suspending the accounts of repeat infringers, and safe harbors for information stored at the direction of users and for information-location tools.69 An affected subscriber may issue a counter-notice, the provider may restore the material if, within seven days of forwarding the counter-notice, it receives no response from the copyright owner indicating that the subscriber lacked authorization, and a dissatisfied party may refer the matter to the Nigerian Copyright Commission, giving the sector regulator a defined adjudicative role without displacing the courts.70 The Act does not state the standard to apply where a notice and counter-notice compete. A provider that fails to comply with the takedown and counter-notice requirements is liable for breach of statutory duty and for the infringement to the same extent as the person who placed the content on the network (section 55(6)).71
South Africa. Section 77 of the Electronic Communications and Transactions Act 2002 establishes a take-down procedure built around notification, without a notice-and-counter-notice structure.72 Because the Act provides no right of reply for the third party whose content is removed, a take-down may occur on an unverified allegation, with the affected party’s only recourse being the ordinary courts.73
Uganda. The Copyright and Neighbouring Rights (Amendment) Act 2026 was assented to on 29 April 2026 and published on 22 May 2026.74 Section 17 of the Act inserts a section 49A under which the Registrar of Copyright or a copyright owner may issue a notice requesting a service provider to block or take down infringing content; a provider that fails to comply commits an offence and may be ordered to pay compensation; the Act permits regulations on the format and procedure of notices; and the Registrar may not issue a notice unless the owner holds a certificate of registration as proof of ownership.75 The text examined does not establish a counter-notice procedure. On the text examined, it strengthens the owner’s position without a verified safeguard for the uploader, and so illustrates the other side of the balance with which this article is concerned: enforcement powers without a demonstrated route of reply.
Global and supplementary benchmarks. The United States conditions platform immunity on the absence of actual knowledge and of direct financial benefit, and on expeditious removal following a compliant notice, with a counter-notification procedure for affected users.76 The European Union goes further, imposing on certain online content-sharing platforms obligations to seek licenses and to act against infringing content.77 India’s 2012 amendment inserted into its Copyright Act an exception for transient or incidental storage, under which a person who receives a written complaint must refrain from facilitating access for twenty-one days or until a court order is received, showing that such a mechanism can be added by targeted amendment of existing provisions.78 India also illustrates a risk: later intermediary rules impose compliance windows measured in hours, a standard demanding for smaller domestic intermediaries.79 Ghana’s Act establishes a copyright monitoring team charged with anti-piracy activity but prescribes no defined takedown procedure, which illustrates that administrative breadth is not a substitute for procedural specificity.80
Findings
The analysis yields five principal findings.
First, the framework’s rights are adequate in form, but its statutory design for online enforcement is incomplete. Sections 9(1) and 45 of the Act are broad enough to capture the acts through which online infringement occurs.81 But the Act supplies no procedural apparatus for vindicating those rights against the actors who make online infringement possible. The Act does not establish a dedicated notice-and-takedown and intermediary-liability procedure. Ordinary civil remedies may remain available against a platform where the facts establish liability under existing law, but the obligations that the platform that hosted, indexed and monetized the file bears under instruments outside the Act were not designed for copyright, and a right whose enforcement depends in practice on identifying and suing individual uploaders is of limited practical value however broadly it is worded.
Secondly, no external instrument compels reform. Tanzania is bound by Berne and TRIPS, neither of which addresses platform-mediated exploitation. It is not a party to the WIPO Copyright Treaty whose language its statute echoes; ARIPO has adopted no protocol on online copyright enforcement; the East African Community has no binding digital copyright instrument; and the AfCFTA Protocol is not in force. Dualism means that even accession would not suffice without amendment of the Act.
Thirdly, the courts can make the rights work, but the civil route is mismatched to online infringement and the administrative route is not designed for copyright. Multichoice proceeded because a single commercial defendant could be identified and served; the Mwinyijuma litigation consumed thirteen years before a jurisdictional preliminary was settled.82 A rights holder who proceeds under the Act must instruct counsel, file a civil suit in the competent court and await service, hearing and judgment, a process whose usual duration cannot be measured on the data available. The notification route under section 45 of the Cybercrimes Act and regulation 7 of the 2016 Regulations offers a far shorter timetable of twenty-four hours and requires neither counsel nor proceedings. But it is a general route: where the provider is not ‘satisfied’ it gives the rights holder only a statement of reasons under regulation 7(2) and a report to a competent authority under section 45(5), neither tied to a copyright-specific standard, and it gives the user no safeguard if the provider is satisfied, and the consequences it attaches to a non-complying provider are criminal, enforced by prosecutors and not by rights holders, so that it gives a rights holder no civil lever against the provider, although the Act’s civil remedies against the uploader remain available. The contrast between a full civil suit and a notification measured in hours is illustrative, not an empirical measurement; the thirteen-year Mwinyijuma history shows how long a single dispute can take but is not evidence of the ordinary duration of copyright litigation, for which no comprehensive data exist; and no published record shows how often the notification route has in fact been used for copyright. The contrast nonetheless explains why doctrinal adequacy and practical adequacy are different questions, and why the existence of a route is not the same as the existence of a remedy.
Fourthly, the statutory allocation of institutional roles is incomplete for platform-mediated infringement, which is a matter of legal design and placement rather than capacity alone. COSOTA’s functions of administration, royalty collection where collective management organizations do not operate and enforcement support are substantial,83 but the Act gives it no power to compel a platform to remove content, to issue a binding takedown order, or to pursue compliance from an intermediary outside Tanzania, and the cybercrime instruments give it no express role. Nothing prevents COSOTA, like any person, from lodging a notification under section 45 on behalf of the rights holders it serves. But no instrument examined here makes it the competent authority to which an ignored notification may be reported, requires a provider to treat its notice differently from anyone else’s, or gives it a function in verifying notices, transmitting them, receiving and evaluating a counter-notice or keeping records. The legislation examined does not expressly assign COSOTA these functions for online infringement, and no published evidence was identified that it operates such a procedure. The inference that this is because no mandate, procedure or resourcing has been directed towards building them is offered as the best available explanation, not as an established fact. Fragmentation compounds the problem: TCRA’s content-removal powers are framed around prohibited content generally; the cybercrime route is framed in the language of offences and law enforcement; and neither coordinates with the Act or with the monitoring and filtering duties that the Online Content Regulations impose. The ordinary civil process, with no expedited track for modest but numerous online complaints, adds procedural rigidity. No published record of online copyright complaints was identified in Tanzania.
Fifthly, the comparative evidence corroborates the mechanism-centered diagnosis. Kenya, Nigeria and Uganda each address intermediary liability for copyright in the copyright statute itself. Kenya and Nigeria pair the notice with a counter-notice, Kenya’s being thin since the provider must disable access unless one is received and the Act does not say what follows, and both attach civil liability to a provider’s failure to comply with a notice, which is the consequence the Tanzanian route lacks. Uganda’s 2026 amendment, on the text examined, provides for notices issued by the Registrar or the owner, an offence and compensation for a non-complying provider, but no verified counter-notice, so that it supplies the enforcement side of the balance without a demonstrated safeguard for users; South Africa relies on a general take-down provision of its electronic communications statute, and its experience shows the cost of omitting a right of reply. Measured against the five comparative parameters, Tanzania meets the first two in part, through the general cybercrime route, but not the third, because no counter-notice or appeal safeguard protects users; not the fourth, because no institutional locus for copyright notices and disputes is defined; and not the fifth, because no copyright-specific cross-border notice-and-takedown or compliance arrangement was identified. The comparison therefore shows that the question for Tanzania is not whether to legislate on intermediaries, which it has done, but how and where.
Discussion of findings
A. From a capacity-centered to a mechanism-centered diagnosis
The most important implication of the findings concerns the order of reform. Magessa, the most recent Tanzanian contribution reviewed in Part IV, frames the enforcement problem substantially as one of institutional capacity: agencies that are under-resourced, under-trained and under-equipped to keep pace with online infringement.84 This article does not dispute that capacity constraints exist. It departs from the capacity-centered account in locating the principal difficulty elsewhere, in the design and placement of the legal mechanisms. Additional funding or training would not give COSOTA a defined place in a notice process that no instrument assigns to it, would not give users a counter-notice that no instrument provides, and would not give providers an objective standard that no instrument states. A reform program that invests in forensic tools and staff training without first creating the legal mechanisms those tools would serve risks equipping an institution to perform functions that no instrument assigns to it.
B. Rejecting the counter-hypothesis of a deliberate choice to rely on general law
A plausible rival explanation is that the absence of a copyright-specific regime reflects a deliberate legislative choice to leave online copyright enforcement to the general cybercrime and online-content instruments, on the view that a specialized regime is unnecessary or premature for the Tanzanian market. Two considerations weigh against it. First, no legislative record, explanatory memorandum or committee report was identified indicating that the omission of an intermediary regime from the Act was considered and adopted as a policy of reliance on those instruments, as distinct from simply pre-dating them or being overtaken by them. Secondly, and more decisively, the general route lacks the safeguards that the Kenyan and Nigerian regimes have in common: a counter-notice procedure and a civil consequence for a provider that ignores a notice. A considered choice to rely on the general route would have had to accept both omissions for copyright, in a market where platforms are central to civic discourse. An unexamined carry-over of provisions drafted for other kinds of unlawful content explains that pattern more plausibly. The caveat that the absence of a record is not proof of the absence of deliberation is acknowledged; the inference is one of best explanation.
C. Applying the dual standard
The enforceability thesis and the least-cost-avoider rationale together support intermediary-directed reform. Because the platform alone can act at scale and low cost, it is the natural addressee of a conditional obligation. But the democratic-culture constraint shapes how that obligation should be framed. The comparative record shows both sides of the point. South Africa’s notification procedure, without a counter-notice, generates over-removal and due-process concerns; Kenya’s 2021 controversy shows that a regime perceived as under-protective of expression will be contested; and the unresolved counter-notice standards in Kenya and Nigeria show that an express regime resolves the threshold problem of silence without resolving every question of calibration. Uganda’s 2026 amendment, on the text examined, provides for notices and provider consequences without a verified counter-notice, which illustrates the opposite risk: that the power to obtain takedown may be strengthened before the safeguard for users is designed. For Tanzania, where platforms are central to civic discourse, a counter-notice and restoration procedure is therefore not an optional refinement but a condition of a regime that satisfies the dual standard. The existing Tanzanian route does not meet that condition on its face: it pairs a short removal period with no right of reply and, through regulations 9 and 15A of the Online Content Regulations, an expectation of monitoring and filtering, which is the form of over-removal risk that the democratic-culture perspective warns against.
D. Limits, qualifications and the regional dimension
Several qualifications temper the analysis. A notice-and-takedown mechanism, however fast, is best suited to preventing the continuation and recurrence of infringement, including re-uploading of removed content, rather than preventing a single live transmission from occurring. For live-streamed events the more relevant remedy is likely to be a standing removal arrangement of the kind large rights holders negotiate directly with major platforms, supplemented by expedited interim injunctive relief. The regime would also fall unevenly: dominant hosting platforms are foreign-domiciled and not directly subject to Tanzanian jurisdiction, which is why the cross-border element cannot be dropped, and why this article treats the practical reach of a domestic regime over foreign platforms as an open question for further research. The Act’s exceptions regime has likewise not been tested against the three-step test in a digital context, and its capacity to accommodate transformative online uses such as parody and remix is untested rather than confirmed adequate. And the burden of the present framework falls hardest on small and independent creators, for whom a civil remedy that requires counsel is illusory. Section 45 and regulation 6 already allow a rights holder to initiate a notice without counsel, and that feature should be preserved; the difficulty is that a notice that can be lodged easily but is answered without a standard, a safeguard or a civil consequence serves small creators poorly when the provider declines to act.
The regional dimension reinforces the urgency. Tanzania is a partner state in a Community whose treaty commits its members to harmonize their policies on the promotion and protection of intellectual property rights,85 yet two of its fellow partner states have legislated for intermediary obligations in their copyright statutes, Kenya with a counter-notice and civil liability for non-compliance and Uganda with notices by the Registrar or owner and offences for non-compliant providers, while Tanzania relies on general instruments that lack a copyright-specific procedure. The comparative lesson is accordingly one of calibration and placement, not transplantation. Four lessons stand out. Jurisdictions that closed a comparable gap did so through a conditional safe harbor rather than strict liability or complete immunity. Every response timeframe in the Kenyan and Nigerian regimes was paired with provision for a counter-notice, which is the pairing that the Tanzanian twenty-four-hour rule lacks. A response period should be set by reference to the procedural steps it must accommodate and to domestic capacity, not by adopting the shortest period found in comparative practice. And a regime that assigns the notice process to no defined institution, as the present Tanzanian regime does, leaves the rights holder without a single door through which to seek relief, so any reform should designate one.
E. The response period: why seventy-two hours
The reform proposed in Part VIII sets a default period of seventy-two consecutive clock hours for the first stage of the procedure, in which a provider must screen a compliant notice, notify the uploader and take initial action by removing or disabling access, which is three times the twenty-four hours that regulation 7 of the Cybercrimes (General) Regulations already prescribes.86 A period longer than the existing one needs justification in an article that argues for stronger enforcement, and the justification is not the arithmetic of the multiple. A period of seventy-two hours is not better than one of sixty or ninety merely because it is three times twenty-four; the multiple is a consequence of the considerations that follow, not a principle from which they derive. Regulation 11(3) and (4) of the Online Content Regulations separately requires a licensee ordered by the Authority or notified by a person affected by the content or existence of prohibited content to notify its subscribers within two hours to remove it, and to suspend or terminate the account of a subscriber who does not remove it within two hours of that notification.87
Four considerations support the figure. First, the twenty-four hours in regulation 7 is firmer in number than in effect. It binds a provider only where it is ‘satisfied’ with the notification. A provider that declines must give reasons (regulation 7(2)) and a complainant may report the failure to a competent authority (section 45(5)), so the route is not without a means of challenge, but neither supplies a copyright-specific standard against which the decision is to be judged. A longer period that is paired with an objective standard for removal, a duty to give reasons for refusal and a civil consequence for non-compliance would give a rights holder a firmer basis than a shorter period that turns on the provider’s own satisfaction. Secondly, the first stage asks a provider to do more than regulation 7 does. It must screen the notice against the objective standard, notify the uploader no later than the moment it removes or disables access, and remove or disable access within the period; removal does not wait for a counter-notice. Regulation 7 compresses the provider’s task into a single step because it has no counter-notice, and a regime that adds the safeguard that the democratic-culture perspective requires necessarily adds time. The seventy-two hours covers only those three actions: lodging, forwarding and deciding a counter-notice run afterwards on the separate timetable in Part VIII and do not extend the period for removal. Thirdly, a period of seventy-two consecutive clock hours, running continuously from receipt of a compliant notice and including weekends and public holidays, gives a provider that receives a notice outside working hours until the evening of the third day: a notice received at 18:00 on a Friday falls due at 18:00 on the Monday, not on the Saturday. That matters for the small domestic providers that lack round-the-clock staffing and answers the concern, noted above in relation to India, that compliance windows measured in hours can be demanding.88 A continuous clock is preferred to a business-hours clock because online infringement does most damage over weekends and in the days after a release, and a business-hours clock would carry the same Friday notice well into the following week. The periods in the counter-notice and restoration timetable are deliberately counted in working days, because those steps are safeguards for the user and the complainant and not deadlines against the provider. Fourthly, the period is an outer limit and a default, not a floor: the proposal preserves the existing twenty-four hours as the shortest period that COSOTA may prescribe by regulation for time-critical categories such as pre-release leaks and live events, so that the speed of regulation 7 is retained where speed matters most. Nigeria’s Copyright Act 2022 offers a statutory comparator for the sequence of notice, notification of the subscriber and restoration following a counter-notice.89
These considerations do not remove two costs, and the article does not claim that they do. A default of seventy-two hours is longer than the period now written into regulation 7, so the reform trades some initial speed in ordinary cases for procedural safeguards and a more structured assessment. And the figure is a reasoned policy judgment, not an empirical finding: no Tanzanian data exist on how long providers take to verify and act on notices. Finally, the proposal must be drafted to displace regulation 7 and the removal duties of the Online Content Regulations for copyright notices. If it were enacted alongside them, rights holders could invoke the shorter periods while providers invoked the longer, and two clocks would run in parallel.
Conclusion and recommendations
A. Conclusion
This article has argued that Tanzania’s framework provides a general legal route for takedown and conditional intermediary protections, but that its statutory design lacks a coherent copyright-specific procedure with clear standards, counter-notice safeguards and an expressly assigned institutional role. The Act’s exclusive rights are broad and technologically neutral, but the Act itself supplies no intermediary-liability regime, no notice-and-takedown procedure and no copyright-specific cross-border mechanism. Tanzanian law does not, however, leave rights holders without a route. Part V of the Cybercrimes Act, 2015 allows any person to notify infringing material (section 45); the Cybercrimes (General) Regulations, 2016 require a response within twenty-four hours (regulation 7); and the Online Content Regulations, 2020 oblige online content hosts to ensure that prohibited contents, which include content infringing intellectual property rights, are removed upon notification by the Authority or affected party (regulation 15), and licensees to monitor and moderate user generated content (regulation 9, as amended in 2022). That route is general and not copyright-specific. It provides no counter-notice procedure, no objective standard for removal and no defined role for the Copyright Office, and the consequences it attaches to a non-complying provider are criminal rather than civil, whereas the Act’s civil remedies run against the infringer and not the intermediary. On the legislation examined, the institutions charged with enforcement are not expressly given the functions that platform-mediated infringement requires, which is a matter of statutory design rather than of capacity alone, and the fragmentation between COSOTA, the communications regulator and the cybercrime regime leaves rights holders without a single door through which to seek relief. The Kenyan and Nigerian regimes contain safeguards that the Tanzanian route lacks, and Uganda’s 2026 amendment shows the opposite risk of strengthening enforcement without a verified safeguard for users; none has been adopted without difficulty. The case for legislative reform is therefore strong, and so is the case for designing it with care: it should build on the existing route, not ignore it.
B. Recommendations
First, enact a copyright-specific conditional safe-harbor and notice-and-takedown regime that builds on the existing route. The Act should be amended by the insertion of a new Part which, for copyright and neighbouring rights, supersedes sections 39 to 45 of the Cybercrimes Act and regulations 6 and 7 of the Cybercrimes (General) Regulations, 2016, so that a single set of rules governs notices of copyright infringement. The drafting should state four things. First, what is amended: the Act gains the new Part; sections 39 to 45 of the Cybercrimes Act and regulations 6 and 7 of the 2016 Regulations are amended to provide that they do not apply to a notice that alleges copyright or neighbouring-rights infringement and is capable of being made under the new Part; and the Online Content Regulations are amended so that regulations 9, 15 and 15A do not, in respect of alleged copyright infringement, impose a monitoring, filtering or removal duty independent of the new Part. Secondly, what remains in force: the Cybercrimes Act and the Online Content Regulations continue to apply to all other unlawful content; the immunities in sections 40 to 44 continue for non-copyright matters; the prohibition of general monitoring in section 39(1) continues; and the offences in the Cybercrimes Act, including section 24, continue to apply to genuinely criminal conduct, so that a copyright notice does not remove a case from the criminal law. Thirdly, how conflicts are resolved: for a notice alleging copyright infringement, the new Part prevails over inconsistent provisions of the Cybercrimes Act, the 2016 Regulations and the Online Content Regulations, and a notice that also alleges non-copyright unlawfulness is governed by the new Part only as to the copyright allegation. Fourthly, a transitional rule: notices lodged before commencement continue under the existing provisions. The new Part should set out, at a minimum, nine elements. (i) A functional definition of ‘online service provider’, covering the provision of transmission, routing or storage of digital content or of a platform through which content is made available to the public, whether or not established in Tanzania, defined by function rather than by a fixed list of platform types so that it accommodates new categories of service. (ii) A conditional exemption from liability for infringement arising solely from third-party content, available where the provider lacks actual knowledge, derives no direct financial benefit where it has the right and ability to control the activity, and expeditiously removes or disables access on receipt of a compliant notice, with an express statement, which should prevail over the monitoring and filtering duties in regulations 9 and 15A of the Online Content Regulations for copyright, that no general monitoring obligation is imposed. (iii) A prescribed notice in writing, which may be given electronically, building on the form in the Schedule to the 2016 Regulations and on section 45(2) of the Cybercrimes Act, and identifying the work, the infringing material with sufficient information to locate it, the complainant’s contact details, and a good-faith statement of non-authorization. (iv) Objective standards for removal in place of the provider’s own satisfaction: a notice is compliant if it contains the prescribed particulars; a provider must act where a compliant notice identifies material that reproduces the whole or a substantial part of the work and the complainant declares ownership and non-authorization; and a provider must decline, and give reasons, where the notice is incomplete or discloses on its face an exception or authorization. (v) A first-stage deadline of seventy-two consecutive clock hours, running continuously from receipt of a compliant notice (recorded by the provider on receipt) and including weekends and public holidays, for the reasons given in Part VII. Within that period the provider must screen the notice against the standard in element (iv); notify the uploader, no later than the moment access is disabled, with a copy of the notice and a statement of the counter-notice procedure; and, unless the notice fails the standard, remove or disable access. COSOTA may prescribe by regulation a shorter period, not less than twenty-four hours, for specified categories of content or provider. (vi) A separate counter-notice and restoration timetable, which starts only after the first stage and does not extend it. The uploader may lodge a counter-notice within ten working days of notification; the provider must forward it to the complainant within twenty-four hours of receipt; and the complainant has ten working days from forwarding to notify the provider and COSOTA that proceedings or a referral to COSOTA have been instituted. If no such notice is received, the provider must restore the content within two working days after that period expires; if it is received, the content remains disabled pending determination. (vii) A statutory role for the Copyright Office, namely COSOTA, as the competent authority for copyright notices: receiving reports of non-compliance, reviewing contested notices on referral, issuing guidance on the objective standards and keeping records, subject to judicial review. (viii) A dedicated civil compliance and liability framework for copyright notices, which would sit alongside the existing offences and not remove the Cybercrimes Act’s function of creating them: a provider that does not comply loses its exemption and is exposed to civil liability under the Act, including injunctions and damages; a person who knowingly submits a false notice is liable in damages; the criminal consequences in section 45(3) and (4) of the Cybercrimes Act are not applied to copyright notices handled under the new Part; and section 24 is reviewed so that criminal liability is reserved for wilful infringement on a commercial scale. The reason for that review is proportionality: section 24 prescribes minimum fines or terms of imprisonment, and a regime that offers civil remedies and a counter-notice for ordinary online infringement should not leave an unqualified minimum penalty for the same conduct. The review needs careful drafting, so that it narrows the penalty and not the reach of the offence over wilful commercial piracy, and it should be undertaken separately from, and after, the notice regime so that the safeguards are not mistaken for a decriminalization of piracy. (ix) Provision for cross-border cooperation, including mutual recognition of takedown notices issued under substantially equivalent procedures and the exchange of information needed to identify persons responsible for content hosted abroad. This design follows the comparative evidence while avoiding the best-efforts prevention obligations of the European Union’s Directive, which presuppose platform compliance infrastructure not yet generally present in Tanzania’s market.90
Secondly, supplement the civil process with an expedited track and align adjacent legislation. The enforcement provisions of the Act should be amended to empower COSOTA, subject to subsequent judicial confirmation, to issue a provisional takedown directive in clear-cut cases of online infringement, in the manner in which the Online Content Regulations already permit TCRA to order the removal of prohibited content.91 The Act’s definition of communication to the public should be extended expressly to platform-based dissemination. The Cybercrimes Act and its Regulations should be amended so that copyright notices are channelled into the new Part, and the disclosure and preservation powers that they already give competent authorities in the investigation of offences should be made available, on court order, to rights holders in civil proceedings to identify the persons responsible for infringing content.92
Thirdly, establish a single point of entry through statutory institutional coordination. The new Part should empower the Minister, on the recommendation of COSOTA and in consultation with TCRA, to make regulations establishing a joint coordination mechanism between the two bodies and linking the National Arts Council, so that a rights holder has one clearly designated entry point. Pending enactment, the Minister should confirm by regulation that COSOTA is a competent authority for copyright notices under the existing Cybercrimes (General) Regulations, which would give section 45(5) of the Cybercrimes Act practical content for copyright at once. COSOTA should develop an online complaint portal for direct submission of notices, and should publish annual aggregate data on notices received and their disposition, a record that would also make future empirical evaluation possible. The judiciary should designate judicial officers within the courts that hear copyright claims for case-management training in copyright and digital intellectual property matters and consider, in the longer term, a specialized commercial or intellectual property division.
Fourthly, sequence implementation and calibrate to capacity. Implementation should proceed in three phases: legislative and regulatory groundwork in the first year, institutional rollout including the coordination mechanism and complaint portal in the second, and judicial capacity-building and international alignment in the third. Any repeat-infringer account-suspension mechanism of the Nigerian type should be subject to prior judicial or tribunal confirmation, to avoid the due-process controversy that attended Kenya’s 2021 amendment bill. Monitoring indicators should include the average time taken to act on a notice, the number and success rate of counter-notices, and the proportion of disputes resolved at agency level. Dedicated resourcing is essential, and technical assistance of the kind contemplated by the WIPO Development Agenda offers a plausible external avenue, but it should be sequenced alongside, not awaited as a precondition for, domestic reform.
Fifthly, pursue international and regional alignment as a secondary measure. Tanzania should accede to the WIPO Copyright Treaty and the WIPO Performances and Phonograms Treaty, ratify the AfCFTA Protocol on Intellectual Property Rights, and use its Community and ARIPO memberships to advocate for binding copyright harmonization and an extension of ARIPO’s work to online copyright enforcement.93 Because of Tanzania’s dualism, accession must be sequenced together with the domestic amendments above; on its own it would have no domestic legal effect.
Three anticipated objections deserve brief response. It may be said that a safe harbor diminishes rights-holder protection by immunizing intermediaries. It does not: immunity is conditional on compliance with the takedown obligation, so a provider that does not cooperate remains exposed to liability under the existing provisions of the Act. It may also be said that a mandatory timeframe will induce over-removal. The counter-notice and restoration procedure and the objective standard are designed to mitigate that risk, and the default seventy-two-hour period is set, for the reasons given in Part VII, as an outer limit that allows for screening and for notice to the uploader. And it may be said that COSOTA lacks the capacity to administer the mechanism. That objection is not dismissed, which is precisely why the recommended sequence builds the legal groundwork a year before institutional rollout, so that capacity is developed deliberately rather than assumed.
*****
Footnotes
1. William R. Cornish, David Llewelyn & Tanya Aplin, Intellectual Property: Patents, Copyright, Trade Marks and Allied Rights 247 (7th ed. 2010).
2. See generally Augustino Magessa, Reassessing Copyright Protection in the Age of Artificial Intelligence: Legal and Policy Challenges in Tanzania, 8 E. Afr. J.L. & Ethics 34, 35, 39 (2025), https://doi.org/10.37284/eajle.8.2.4005.
3. Copyright and Neighbouring Rights Act, Cap. 218 R.E. 2023, §§ 4, 9(1) (Tanz.) [hereinafter Copyright Act] (defining “communication to the public” to include access “at different places and or times individually chosen by them”).
4. Copyright Act, supra note 3, § 45.
5. See generally Copyright Act, supra note 3, pts. V–VI, §§ 39–48 (providing civil remedies, offences and conservatory measures, but no provision on intermediaries, takedown or cross-border enforcement).
6. Cybercrimes Act, No. 14 of 2015, §§ 24, 39–45 (Tanz.) (now Cap. 443 R.E. 2023) [hereinafter Cybercrimes Act].
7. Cybercrimes (General) Regulations, 2016, GN No. 224 of 2016, regs. 6–7 & sched. (Tanz.) [hereinafter Cybercrimes Regulations].
8. Electronic and Postal Communications (Online Content) Regulations, 2020, GN No. 538 of 2020, regs. 9, 15 & sched. 3, para. 6 (Tanz.), as amended by Electronic and Postal Communications (Online Content) (Amendment) Regulations, 2022, GN No. 136 of 2022, reg. 8 (Tanz.) (substituting reg. 9(c)), and Electronic and Postal Communications (Online Content) (Amendment) Regulations, 2025, GN No. 57 of 2025 (Tanz.) (inserting reg. 15A) [hereinafter Online Content Regulations].
9. 17 U.S.C. § 512 (enacted by the Digital Millennium Copyright Act, Pub. L. No. 105-304, 112 Stat. 2860 (1998)).
10. Copyright Act, No. 12 of 2001, Cap. 130, §§ 35A–35D (Kenya) [hereinafter Kenya Copyright Act], inserted by Copyright (Amendment) Act, No. 20 of 2019 (Kenya).
11. Copyright Act, 2022, No. 8 of 2023, pt. VII (Nigeria) [hereinafter Nigeria Copyright Act].
12. Copyright and Neighbouring Rights (Amendment) Act, Act 8 of 2026, § 17 (Uganda) [hereinafter Uganda Amendment Act] (assented to Apr. 29, 2026 and published May 22, 2026) (inserting § 49A into the Copyright and Neighbouring Rights Act, Cap. 222).
13. Multichoice Tanzania Ltd. v. Kiganza, Civil Appeal No. 166 of 2020, [2021] TZHC 4208 (High Ct. Dar es Salaam July 2, 2021) (Tanz.), https://www.wipo.int/wipolex/en/text/587917.
14. Mwinyijuma v. MIC Tanzania Ltd., Civil Case No. 17 of 2012 (Ilala Dist. Ct. Apr. 11, 2016) (Tanz.), rev’d, MIC Tanzania Ltd. v. Mwinyijuma, Civil Appeal No. 112 of 2019 (High Ct. Dar es Salaam Nov. 20, 2020) (Tanz.), rev’d and remanded, Mwinyijuma v. Honora Tanzania PLC, Civil Appeal No. 42 of 2023, [2025] TZCA 822 (Ct. App. Dar es Salaam Aug. 7, 2025) (Tanz.), https://www.wipo.int/wipolex/en/judgments/details/3110.
15. Mark Rose, Authors and Owners: The Invention of Copyright 10–25 (1993).
16. See generally David Bainbridge, Intellectual Property (8th ed. 2010).
17. Paul Goldstein & P. Bernt Hugenholtz, International Copyright: Principles, Law, and Practice 67–72 (4th ed. 2019).
18. Copyright Act, supra note 3, §§ 39–47.
19. 17 U.S.C. § 512(c), (g).
20. WIPO Copyright Treaty arts. 6, 8, Dec. 20, 1996 (entered into force Mar. 6, 2002) [hereinafter WCT].
21. World Intell. Prop. Org., WIPO Copyright Treaty: Contracting Parties, WIPO Lex, https://www.wipo.int/wipolex/en/treaties/ShowResults?search_what=C&treaty_id=16 (last visited Nov. 5, 2026) [hereinafter WCT Contracting Parties].
22. Kenya Copyright Act, supra note 10, §§ 35A–35D.
23. Nigeria Copyright Act, supra note 11, pt. VII.
24. Uganda Amendment Act, supra note 12, § 17 (inserting § 49A).
25. William Fisher, Theories of Intellectual Property, in New Essays in the Legal and Political Theory of Property 168, 171–75 (Stephen R. Munzer ed., 2001).
26. William M. Landes & Richard A. Posner, An Economic Analysis of Copyright Law, 18 J. Legal Stud. 325, 328–33 (1989).
27. Landes & Posner, supra note 26, at 333–36.
28. Guido Calabresi, The Costs of Accidents: A Legal and Economic Analysis 135–40 (1970).
29. William Landes & Douglas Lichtman, Indirect Liability for Copyright Infringement: Napster and Beyond, 17 J. Econ. Persp. 113, 114–16 (2003).
30. See Religious Tech. Ctr. v. Netcom On-Line Commc’n Servs., Inc., 907 F. Supp. 1361 (N.D. Cal. 1995).
31. Neil Weinstock Netanel, Copyright and a Democratic Civil Society, 106 Yale L.J. 283, 288 (1996).
32. Netanel, supra note 31, at 347–50.
33. Jennifer M. Urban & Laura Quilter, Efficient Process or “Chilling Effects”? Takedown Notices Under Section 512 of the Digital Millennium Copyright Act, 22 Santa Clara Computer & High Tech. L.J. 621, 625–28 (2006).
34. Cornish, Llewelyn & Aplin, supra note 1, at 247.
35. Bainbridge, supra note 16; Rose, supra note 15, at 10–25.
36. Goldstein & Hugenholtz, supra note 17, at 51–55.
37. Pamela Samuelson, Digital Media and the Changing Face of Intellectual Property Law, 16 Rutgers Computer & Tech. L.J. 323, 330–35 (1990).
38. Magessa, supra note 2, at 42–44.
39. Urban & Quilter, supra note 33, at 625–28.
40. Berne Convention for the Protection of Literary and Artistic Works, Sept. 9, 1886, as revised at Paris July 24, 1971, 1161 U.N.T.S. 3.
41. Agreement on Trade-Related Aspects of Intellectual Property Rights, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, 1869 U.N.T.S. 299 [hereinafter TRIPS Agreement].
42. TRIPS Agreement, supra note 41, art. 41(1).
43. WCT, supra note 20, arts. 6–8; WIPO Performances and Phonograms Treaty, Dec. 20, 1996 (entered into force May 20, 2002) [hereinafter WPPT].
44. WCT Contracting Parties, supra note 21; World Intell. Prop. Org., Member Profile: United Republic of Tanzania, WIPO Lex, https://www.wipo.int/wipolex/en/members/profile/TZ (last visited Nov. 5, 2026).
45. Tanz. Const. art. 63(3)(d)–(e) (1977) (empowering Parliament to enact legislation where implementation requires it and to ratify treaties).
46. African Reg’l Intell. Prop. Org., Member States, https://www.aripo.org/member-states (last visited Nov. 5, 2026); Lusaka Agreement on the Creation of the African Regional Intellectual Property Organization, Dec. 9, 1976 [hereinafter Lusaka Agreement].
47. Swakopmund Protocol on the Protection of Traditional Knowledge and Expressions of Folklore Within the Framework of the African Regional Intellectual Property Organization, Aug. 9, 2010; Kampala Protocol on Voluntary Registration of Copyright and Related Rights Within the Framework of the African Regional Intellectual Property Organization art. 2, Aug. 28, 2021, https://www.aripo.org/storage/resources-copyright-related-rights-treaties/1674824283_phpjiukhU.pdf.
48. Treaty for the Establishment of the East African Community art. 103(1)(i), Nov. 30, 1999 (as amended) [hereinafter EAC Treaty].
49. Trade Law Centre, Status of AfCFTA Ratification (Sept. 7, 2026), https://www.tralac.org/resources/infographic/13795-status-of-afcfta-ratification.html (last visited Oct. 27, 2026).
50. Protocol to the Agreement Establishing the African Continental Free Trade Area on Intellectual Property Rights arts. 11, 26(d), Feb. 19, 2023 [hereinafter AfCFTA IP Protocol].
51. AfCFTA IP Protocol, supra note 50, art. 34(2); Trade Law Centre, AfCFTA Protocol on Intellectual Property Rights: Factsheet, https://www.tralac.org/resources/infographic/16151-afcfta-protocol-on-intellectual-property-rights-factsheet.html (last visited June 20, 2026).
52. Tanz. Const. art. 24(1); cf. Att’y Gen. v. Akonaay, [1995] TLR 80 (CA) (Tanz.) (holding customary rights in land to be property protected by article 24).
53. Copyright Act, supra note 3, § 4.
54. Copyright Act, supra note 3, §§ 39–47, 51.
55. Mwinyijuma v. Honora Tanzania PLC, supra note 14, at 10–14 (construing section 4 as it stood before amendment); Copyright Act, supra note 3, § 4 (defining “court” as “the court of competent jurisdiction”), as amended by Act No. 9 of 2019, § 12 (Tanz.).
56. Cybercrimes Act, supra note 6, §§ 24, 39–41, 45.
57. Cybercrimes Regulations, supra note 7, regs. 2, 6–7 & sched.
58. Online Content Regulations, supra note 8, regs. 9, 15, 15A, 19 (made under the Electronic and Postal Communications Act, Cap. 306 R.E. 2023 (Tanz.)).
59. Electronic and Postal Communications (Online Content) (Amendment) Regulations, 2026, GN No. 158i of 2026 (Tanz.), https://www.tcra.go.tz/laws-and-guidelines/regulations/the-electronic-and-postal-communications-online-content-amendment-regulations-20-2 (last visited Nov. 5, 2026).
60. National Arts Council Act, Cap. 204 R.E. 2002 (Tanz.).
61. Multichoice, supra note 13.
62. Mwinyijuma v. Honora Tanzania PLC, supra note 14, at 1–3, 15.
63. Kenya Copyright Act, supra note 10, §§ 35A–35D, 35A(1)(c)(iii); Digital Copyright Protection: Internet Service Providers Under the Copyright (Amendment) Act of 2019, MMAN Advocates (Dec. 19, 2019), https://mman.co.ke/content/digital-copyright-protection-internet-service-providers-under-copyright-amendment-act-2019 [hereinafter MMAN Commentary]; Caroline Wanjiru, Internet Service Providers (ISPs) and Copyright in Kenya: Commentary on the Copyright Amendment Act 2019, CIPIT (Oct. 18, 2019), https://cipit.strathmore.edu/internet-service-providers-isps-and-copyright-in-kenya-commentary-on-the-copyright-amendment-act-2019/.
64. Kenya Copyright Act, supra note 10, § 35B(4)–(6).
65. Kenya Copyright Act, supra note 10, §§ 35B(2), 35B(8), 35C(1)(a).
66. Kenya: Copyright Bill Must Respect International Standards of Free Speech, Article 19 (Jan. 12, 2022), https://www.article19.org/resources/kenya-proposed-copyright-amendment-bill-must-respect-free-speech/; Kenya: 2021 Changes to Copyright Law, Spoor & Fisher (Jan. 25, 2022), https://spoor.com/kenya-2021-changes-to-copyright-law.
67. Copyright Act, Cap. 130, §§ 35A–35D (Rev. 2022) (Kenya), https://new.kenyalaw.org/akn/ke/act/2001/12/eng@2022-12-31.
68. MMAN Commentary, supra note 63.
69. Nigeria Copyright Act, supra note 11, pt. VII, §§ 54–59.
70. Nigeria Copyright Act, supra note 11, § 55(2), (4); Deborah Dada, Liability of Internet Service Providers (ISPs) in Nigeria: Understanding the Safe Harbour Provisions, Mondaq (Apr. 26, 2024), https://www.mondaq.com/nigeria/copyright/1456386.
71. Nigeria Copyright Act, supra note 11, § 55(6); see also Mosun Oke, Tolulope Oguntade & Aramide Oladimeji, Online Intermediary Liability in Nigeria: What Platforms Need to Know, Tope Adebayo LP (Oct. 9, 2026), https://topeadebayolp.com/online-intermediary-liability-in-nigeria-what-platforms-need-to-know/.
72. Electronic Communications and Transactions Act 25 of 2002 § 77 (S. Afr.).
73. Alex Comninos, Intermediary Liability in South Africa 12–13 (Ass’n for Progressive Commc’ns, Intermediary Liability in Africa Research Paper No. 3, 2012), https://www.apc.org/sites/default/files/Intermediary_Liability_in_South_Africa-Comninos_06.12.12_0.pdf.
74. Uganda Amendment Act, supra note 12.
75. Uganda Amendment Act, supra note 12, § 17 (inserting § 49A); Museveni Signs Copyright Law, Seven Others, Daily Monitor (Apr. 29, 2026), https://www.monitor.co.ug/uganda/news/national/museveni-signs-copyright-law-seven-others-5441370.
76. 17 U.S.C. § 512(c)(1), (g)(2)–(3).
77. Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on Copyright and Related Rights in the Digital Single Market, art. 17, 2019 O.J. (L 130) 92.
78. The Copyright Act, 1957, § 52(1)(c), No. 14, Acts of Parliament, 1957 (India), as amended by The Copyright (Amendment) Act, 2012, No. 27, Acts of Parliament, 2012 (India).
79. Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, r. 3(1)(d) (India), as amended by G.S.R. 120(E) (Feb. 10, 2026) [hereinafter Intermediary Rules].
80. Copyright Act, 2005, Act 690, § 50 (Ghana).
81. Copyright Act, supra note 3, §§ 9(1), 45.
82. Multichoice, supra note 13; Mwinyijuma v. Honora Tanzania PLC, supra note 14.
83. Copyright Act, supra note 3, § 51.
84. Magessa, supra note 2, at 43–44.
85. EAC Treaty, supra note 48, art. 103(1)(i).
86. Cybercrimes Regulations, supra note 7, reg. 7.
87. Online Content Regulations, supra note 8, reg. 11(3)–(4).
88. Intermediary Rules, supra note 79, r. 3(1)(d).
89. Nigeria Copyright Act, supra note 11, §§ 54–55.
90. Directive (EU) 2019/790, supra note 77, art. 17(4), (8).
91. Online Content Regulations, supra note 8, regs. 9(f), 19(b).
92. Cybercrimes Act, supra note 6, §§ 32–35; Cybercrimes Regulations, supra note 7, regs. 3–4.
93. WCT, supra note 20; WPPT, supra note 43; AfCFTA IP Protocol, supra note 50; Lusaka Agreement, supra note 46, art. III.