Artificial intelligence (AI) is reshaping how governments generate information, analyse complex problems and support strategic decisions (Caiza et al., 2024). Unlike earlier digital government that mainly automated administrative processes, contemporary AI can identify patterns, generate predictions, synthesise datasets and provide recommendations that influence policy and organisational choices. The OECD (2025) reports that governments use AI for decision-making, sense-making and forecasting, while emphasising governance, data quality, skills and human oversight. This development shifts attention towards the relationship between AI-generated intelligence and human actors who interpret and use it.
Furthermore, the United Arab Emirates (UAE) provides a setting because its digital-government agenda places AI at the centre of public-sector development. The UAE Strategy for Artificial Intelligence and digital-government initiatives seek to enhance government performance, service delivery and data-driven decision-making. Recent UAE evidence indicates that AI is used to improve efficiency and decision-making, although challenges involving data, privacy, security, infrastructure and user capabilities remain (Akhoirshieda et al., 2024). The OECD (2025) similarly stresses opportunities for better decision-making alongside risks involving transparency, accountability and excessive reliance on automated outputs. The UAE is therefore an important context for examining how leaders engage with AI-generated intelligence.
The literature increasingly recognises that AI-supported decision-making depends on how humans interact with, interpret and act upon outputs. Research on human–AI collaboration highlights complementary capabilities, calibrated trust, explanation and allocation of decision tasks (Bao et al., 2023; Raees et al., 2024). Evidence also shows that human involvement does not automatically guarantee better decisions. Sele and Chugunova (2024) found that human-in-the-loop arrangements could increase acceptance of automated recommendations while reducing intervention with inaccurate recommendations. Steyvers and Kumar (2024) identify challenges concerning reliance and human oversight. Strategic decision-making therefore requires attention to how decision-makers accept, question, modify or override AI-generated recommendations.
Despite these developments, a problem remains insufficiently examined: existing research provides limited empirical understanding of how government decision-makers in the UAE negotiate human judgement, expertise and decision authority when working with AI-generated intelligence. Akhoirshieda et al. (2024) identified a shortage of UAE public-sector AI studies, while broader government research indicates that evidence on AI use in policymaking and strategic functions remains limited (OECD, 2025). The gap is therefore not simply whether UAE institutions use AI, but how human–AI collaboration operates within strategic decision processes and how leaders preserve meaningful judgement and authority. This study addresses the gap through a qualitative review of the literature and of UAE policy documents.
Accordingly, the study asks: What is the nature of human–AI collaboration in strategic decision-making within UAE government institutions? How do government decision-makers interpret, evaluate and balance AI-generated recommendations with human judgement and decision authority? In alignment with these questions, the study seeks to examine the nature of human–AI collaboration in strategic decision-making within UAE government institutions and to assess how decision-makers balance AI-generated intelligence with human judgement, expertise and decision authority. The inquiry is confined to UAE government institutions and focuses on strategic decisions involving AI-supported information, recommendations and analysis. Its significance lies in clarifying human processes shaping AI-supported choices, contributing empirical insight to scholarship and informing responsible AI governance and leadership practice.
This study adopts Dynamic Capabilities Theory as its theoretical premise for examining human–AI collaboration, strategic decision-making and the transformation of government leadership within UAE government institutions. The theory was developed by Teece et al. (1997) to explain how organisations adapt to rapidly changing technological and environmental conditions by developing, integrating and reconfiguring internal capabilities. Teece (2007) subsequently elaborated its microfoundations through three interconnected processes: sensing, seizing and reconfiguring. Sensing involves identifying emerging opportunities and threats; seizing concerns mobilising resources and making strategic choices; while reconfiguring involves transforming organisational resources, structures and routines in response to continuing change.
Five key assumptions underpin the analytical application of the theory in this study: organisations operate within changing environments; existing resources and capabilities may become inadequate under technological change; organisations must sense emerging opportunities and threats; they must seize opportunities through strategic decisions and resource mobilisation; and they must reconfigure resources, structures, processes and routines to sustain adaptation. These assumptions distinguish dynamic capabilities from static resource perspectives by emphasising an organisation’s capacity to renew and transform its capabilities rather than merely possessing resources. Teece et al. (1997) and Teece (2007) establish these processes as central mechanisms through which organisations respond strategically to environmental change.
Recent public-sector research supports the relevance of this perspective to AI transformation. Selten and Klievink (2024) found that public organisations adopting AI must reconcile established bureaucratic structures with the flexibility required for technological innovation, demonstrating that AI adoption requires organisational adaptation rather than technological implementation alone.
The researcher adopts Dynamic Capabilities Theory because AI increasingly affects government leadership, organisational processes and strategic decision-making. The assumption concerning reconfiguring capabilities provides the strongest basis for the study’s scientific contribution by extending the theory to examine how AI can transform leadership roles, practices and organisational relationships. The study therefore explains leadership transformation through sensing AI opportunities and risks, seizing AI-enabled possibilities, and reconfiguring leadership and organisational practices within UAE government institutions.
Human–AI collaboration refers to decision processes in which human decision-makers and artificial intelligence systems contribute complementary capabilities to the achievement of a shared task or outcome (Hemmer et al., 2025). The literature has progressively moved beyond the earlier human–computer interaction paradigm, in which technology was largely treated as a tool controlled by the user, towards human-centred and interactive forms of AI in which systems provide predictions, recommendations, explanations and adaptive assistance. Recent scholarship conceptualises this development as human–AI teaming or synergy, emphasising that the relevant unit of analysis is no longer the AI system alone but the interaction between human capabilities, machine capabilities and the task environment (Bao et al., 2023; Raees et al., 2024). This shift is particularly important for strategic decision-making because AI can process large volumes of information and identify patterns at a scale that exceeds ordinary human analytical capacity, while human decision-makers retain contextual, ethical and experiential capabilities that remain difficult to encode computationally.
Accordingly, AI is increasingly understood as a collaborative partner rather than merely a passive decision-support instrument. Effective collaboration depends on the distribution of tasks according to the relative strengths and limitations of humans and machines. Steyvers and Kumar (2024) argue that genuine human–AI complementarity occurs when joint performance exceeds that of either the unaided human or AI operating independently. Such complementarity requires decision-makers to recognise when AI assistance is useful and when human expertise should take precedence. Similarly, recent empirical research on human–generative-AI collaboration suggests that AI can be particularly valuable in information-intensive and unfamiliar tasks, while human contextual reasoning and creativity remain important where decisions require interpretation beyond available data (Hao et al., 2024). Thus, collaboration should not be equated with simple acceptance of machine recommendations. It involves an iterative process of information exchange, evaluation, coordination and adjustment in which responsibility for particular tasks may shift between human and AI actors.
Trust constitutes a central condition of this relationship because decision-makers must determine when AI outputs are sufficiently reliable to warrant reliance. Recent research increasingly distinguishes appropriate reliance from unconditional trust, emphasising the importance of calibrated trust that corresponds to the actual capabilities and limitations of an AI system. Steyvers and Kumar (2024) demonstrate that accurate mental models of AI are essential because users need to understand not only what an AI system can do but also where its performance may be unreliable. More recent work similarly conceptualises trust in human–AI teams as dynamic and influenced by task context, interaction processes and changing perceptions of system performance (Duan et al., 2025). Explainability and interpretability can support this process by helping users understand AI outputs, although their effects are not universally positive. Wang and Ding (2024), for example, found that explanations can improve decision accuracy and behavioural reliance under particular conditions, while other evidence indicates that poorly designed explanations may increase perceived complexity rather than improve understanding. Therefore, transparency should be considered in relation to users’ capabilities and the decision context rather than treated as an automatic solution to trust problems.
A further concern is automation bias, whereby decision-makers may place excessive weight on AI recommendations and insufficiently scrutinise potentially erroneous outputs. Recent scholarship identifies automation bias, inappropriate reliance and failures of human–AI coordination as continuing obstacles to effective collaboration (Romeo & Conti, 2026; Schmutz et al., 2024). Importantly, human–AI interaction can itself influence subsequent human judgement, meaning that AI does not simply support decisions but may alter the cognitive processes through which people evaluate information (Glickman & Sharot, 2025). Effective collaboration therefore requires mechanisms through which decision-makers can question, modify or override AI recommendations rather than merely comply with them. Recent research indicates that decision control can improve users’ perceptions and compliance, whereas explanations alone may have inconsistent effects depending on task complexity and user characteristics (Westphal et al., 2023). Despite these advances, important gaps remain. Much of the empirical literature examines experimental, commercial or specialised task environments, while comparatively less is known about how senior public-sector decision-makers actually negotiate AI recommendations, exercise professional judgement and determine when to accept or override machine-generated advice in complex governmental settings. This gap provides a strong basis for examining human–AI collaboration as a situated decision-making process within UAE government institutions.
Strategic decision-making involves identifying and selecting among longer-term courses of action whose consequences may shape organisational direction and performance. Such decisions are characterised by complexity, ambiguity, uncertainty, delayed feedback and substantial information requirements, which can exceed the cognitive capacity of individual decision-makers (Csaszar et al., 2024). In government institutions, these characteristics are compounded by public accountability, legal obligations, competing policy objectives, resource constraints and diverse stakeholder interests. Consequently, strategic government decisions cannot be reduced to technically optimal solutions because decision-makers must interpret evidence within specific institutional and policy contexts. Recent public-sector research consequently identifies a persistent challenge in aligning technically sophisticated AI models with the realities of governmental decision-making (Fischer-Abaigar et al., 2024).
AI-generated intelligence expands the informational foundation of strategic decision-making by processing large and diverse datasets and producing predictions, patterns, evaluations and recommendations. Its contribution can extend across information search, analysis, forecasting, generation of alternatives and evaluation of strategic options. Csaszar et al. (2024) provide empirical evidence that large language models can generate and evaluate strategic alternatives, demonstrating AI’s potential to augment important cognitive processes underlying strategic decision-making. In government, such capabilities may support data-driven choices by enabling decision-makers to examine extensive information and identify relationships that might otherwise remain difficult to detect. However, AI-generated outputs should be understood as decision-relevant intelligence rather than decisions themselves because their usefulness depends on data quality, model performance and contextual relevance.
AI-supported forecasting and scenario analysis can further broaden the range of alternatives considered by decision-makers, but computational sophistication does not eliminate uncertainty. Fischer-Abaigar et al. (2024) identify misalignment between machine-learning assumptions and public-sector realities as a significant challenge, while the OECD (2025) notes that skewed data, insufficient transparency and overreliance on AI can produce erroneous outcomes and undermine accountability. The reliability of AI-generated intelligence therefore requires critical assessment rather than automatic acceptance. This is particularly important in government, where strategic objectives may incorporate legal, ethical, social and political considerations that cannot always be represented through quantitative variables.
The relationship between AI-generated intelligence and strategic choice can therefore be understood as an interactive process in which AI processes information and generates analytical outputs, while decision-makers interpret, evaluate and contextualise those outputs before selecting a course of action. As synthesised in Table 1, AI contributes primarily through information processing, prediction, scenario analysis and recommendation, whereas strategic choice remains dependent on the quality of the intelligence produced and its contextual interpretation.
Table 1. AI-Generated Intelligence and Its Contribution to Strategic Decision-Making
| Dimension | AI-generated contribution | Implication for strategic decision-making |
|---|---|---|
| Information processing | Analyses large and diverse datasets | Expands the evidence available to decision-makers |
| Prediction | Identifies patterns and estimates possible outcomes | Supports forecasting and anticipation |
| Scenario analysis | Generates or compares possible situations | Broadens consideration of strategic alternatives |
| Recommendation | Produces potential courses of action | Supports evaluation rather than automatically determining choice |
| Information quality | Depends on data, model design and system performance | Requires critical assessment of reliability |
| Contextual interpretation | Processes information within available data and model parameters | Requires human knowledge of institutional and policy circumstances |
| Strategic choice | Generates intelligence and possible alternatives | Requires human evaluation of objectives and consequences |
Source: Author’s own, developed from Csaszar et al. (2024), OECD (2025) and Fischer-Abaigar et al. (2024).
As indicated in Table 1, the contribution of AI is predominantly analytical and informational rather than determinative. Although AI can process extensive information, identify patterns, generate predictions and formulate recommendations, the movement from these outputs to strategic choice requires contextual interpretation and evaluation. This distinction is particularly important in government institutions, where strategic decisions may involve considerations that cannot be fully represented through computational models. The literature therefore establishes an important unresolved question: how do government decision-makers interpret, evaluate and incorporate AI-generated intelligence into strategic choices? Existing research demonstrates AI’s capacity to augment strategic analysis, but comparatively less is known about the human process through which public-sector decision-makers translate AI-generated information, predictions and recommendations into strategic action. This gap provides the basis for examining human judgement, expertise and decision authority in AI-supported decisions.
Human judgement refers to the process through which decision-makers interpret available information, assess alternatives and form evaluations or choices under conditions of uncertainty. In AI-supported environments, judgement remains important because machine-generated recommendations do not necessarily capture contextual, professional or value-based considerations surrounding a decision (Xu et al., 2025). Recent research indicates that effective human–AI decision-making depends on understanding the complementary roles of computational analysis and human judgement rather than assuming that one should replace the other (Srivastava et al., 2025). Professional and managerial expertise can provide contextual knowledge that enables decision-makers to assess whether an AI recommendation is appropriate to the circumstances. Thus, expertise does not simply represent accumulated knowledge; it also influences how decision-makers interpret, question and contextualise AI-generated outputs. Empirical evidence shows that experts may become more sceptical when AI recommendations conflict with their professional judgement, demonstrating that expertise can shape both acceptance and rejection of algorithmic advice.
Human judgement is nevertheless subject to cognitive limitations, including information overload, bounded attention and susceptibility to biases. AI can potentially reduce some of these burdens by processing large quantities of information and identifying patterns that may be difficult for individuals to detect. However, collaboration can also introduce new cognitive risks. Sele and Chugunova (2024), for example, found evidence of automation bias in which participants followed algorithmic recommendations closely and were less likely to correct recommendations containing larger errors. This finding challenges the assumption that keeping a human “in the loop” automatically guarantees effective human oversight. The problem is particularly significant when decision-makers perceive AI outputs as objective or technically superior and consequently reduce the level of independent scrutiny applied to machine recommendations.
Conversely, human decision-makers may resist or discount AI recommendations, particularly when outputs conflict with professional experience, intuition or prior beliefs. Recent evidence demonstrates that experts may accept AI recommendations more readily when they correspond with their initial judgements, while showing greater scepticism when recommendations are inconsistent with professional assessments (Bashkirova & Krpan, 2024). Such findings suggest that neither unconditional reliance nor systematic rejection represents effective human–AI collaboration. Rather, decision quality depends on the ability to distinguish circumstances in which AI provides useful evidence from those in which contextual human knowledge warrants further investigation or intervention. Research on expert–AI pairings similarly shows that professionals actively interpret, negotiate and sometimes resist AI outputs rather than simply following them (Cruz, 2024). This debate places human agency and decision authority at the centre of AI-supported decision-making. Human control requires more than the formal presence of a human decision-maker; it requires meaningful capacity to understand, assess and intervene in AI-supported processes. Tsamados et al. (2025) distinguish supervisory control from human–AI teaming, while Van den Bosch et al. (2025) emphasise meaningful human control as a condition for responsible human–AI collaboration. Decision authority therefore involves the capacity to accept, modify or override AI recommendations when contextual circumstances or professional judgement warrant intervention. Accountability must correspondingly remain identifiable despite the involvement of AI, because delegating analytical tasks to an algorithm does not necessarily transfer responsibility for the resulting decision.
The literature consequently suggests that effective AI-supported strategic decision-making requires a balance between algorithmic capability and human judgement. AI can extend analytical capacity, while human expertise contributes contextual interpretation, ethical reasoning and responsibility. The central unresolved issue is therefore not whether humans or AI should make decisions independently, but how decision-makers determine when to rely on, question, modify or override AI recommendations while retaining meaningful decision authority and accountability. This question is particularly important in government institutions, where strategic decisions may have consequences extending beyond organisational performance to public interests and institutional legitimacy. It provides the foundation for examining how trust, explainability and interaction conditions influence the acceptance or rejection of AI-generated intelligence.
The UAE provides a distinctive context for examining human–AI strategic decision-making because its digital-government development has progressively moved from service digitisation towards data-driven and AI-enabled government. The UAE Digital Government Strategy 2025 places data-driven government, digital-by-design practices, proactive services and digital leadership at the centre of government development, while the UAE Strategy for Artificial Intelligence, launched in 2017, explicitly seeks to enhance government performance through AI and integrate AI into government services and data analysis (UAE Government, 2017, 2024). This policy trajectory establishes an institutional environment in which AI is not simply an optional technological tool but an increasingly important component of government capability. Consequently, the UAE provides an appropriate setting for examining how government leaders and decision-makers interact with AI-generated intelligence when addressing strategic choices.
Recent evidence confirms growing AI adoption across UAE public institutions, although the existing literature remains uneven in its analytical focus. Akhoirshieda et al. (2024), in a systematic review of 20 Scopus-indexed studies, found that AI applications in the UAE public sector are associated with efficiency, service delivery and decision-making, while also identifying continuing concerns relating to data, privacy, security, infrastructure and user-related challenges. More recent UAE research similarly indicates that government employees are increasingly engaging with AI-enabled technologies. Shaer et al. (2025), examining generative AI adoption among Dubai government employees, found evidence of growing use for productivity, task automation and decision support, while Alkaabi et al. (2026) identified perceived usefulness, ease of use and prior AI experience as important factors associated with public-sector AI adoption. These studies demonstrate expanding adoption, but they do not fully explain how AI-generated intelligence is incorporated into strategic human judgement.
The emerging UAE literature also highlights trust, transparency and responsible use as important conditions for AI implementation. Research on ChatGPT integration in the UAE government sector identifies concerns surrounding data accuracy, transparency, privacy, human supervision and accountability (Goher, 2025). Similarly, the UAE’s evolving policy environment emphasises responsible AI governance, while the government’s AI adoption guidance identifies governance, regulation, skills and infrastructure as key enablers of public-sector AI adoption (UAE Government, 2024). However, existing studies tend to examine adoption, acceptance, implementation or governance rather than the micro-level interaction through which decision-makers evaluate, accept, modify or override AI-generated recommendations.
This constitutes an important empirical and theoretical gap. Existing UAE research provides useful evidence about AI adoption and institutional readiness, but comparatively limited attention has been given to the relationship between AI-generated intelligence, human expertise, contextual judgement and strategic decision authority within government institutions. The literature therefore does not sufficiently explain how leaders balance machine recommendations with professional judgement, how trust is calibrated, or how accountability is maintained when AI contributes to consequential strategic choices. The present study addresses this gap by examining these processes through the literature and the UAE’s policy documents on AI in government decision-making. It therefore shifts the analytical focus from whether government institutions adopt AI to how humans and AI interact in the production and evaluation of strategic decisions.
The thematic analysis of the reviewed literature reveals that human–AI collaboration is increasingly becoming an important feature of strategic decision-making within UAE government institutions. References to AI-supported information processing, predictive analysis, pattern recognition, decision recommendations, and adaptive assistance appear consistently across the reviewed studies, demonstrating the growing integration of artificial intelligence into strategic decision processes. However, the analysis simultaneously identifies the continuing importance of human judgement, particularly in contextual interpretation, ethical considerations, professional experience, and the exercise of final decision authority. The findings suggest that effective human–AI collaboration does not involve the passive acceptance of machine-generated recommendations, but rather an interactive process in which AI capabilities and human expertise are combined according to their respective strengths and limitations. At the same time, concerns relating to trust, explainability, automation bias, and inappropriate reliance indicate that the increasing use of AI may require existing decision-making practices and organisational routines to be adjusted.
Consequently, the finding is consistent with existing literature which identifies human–AI complementarity as an important mechanism for improving decision-making in information-intensive environments, while highlighting the continuing importance of human oversight and contextual judgement (Bao et al., 2023; Raees et al., 2024; Steyvers & Kumar, 2024). From the perspective of Dynamic Capabilities Theory, the finding demonstrates the capacity of government institutions to sense emerging opportunities and threats through AI-supported analysis, seize these opportunities through strategic human judgement and resource mobilisation, and reconfigure decision-making routines to accommodate AI-supported processes. Previous studies similarly note that automation bias and inappropriate reliance can undermine effective human–AI collaboration, thereby reinforcing the need for mechanisms that allow decision-makers to scrutinise and override AI recommendations where necessary. Overall, the finding indicates that the strategic value of AI within UAE government institutions depends not merely on its adoption, but on the organisation’s capacity to continuously adapt its human and technological capabilities to changing decision-making environments.
The thematic analysis of the reviewed literature reveals that AI-generated intelligence is increasingly expanding the informational foundation of strategic decision-making within UAE government institutions. The findings indicate that AI contributes to information processing, prediction, pattern identification, scenario analysis, evaluation of alternatives, and generation of strategic recommendations, thereby enabling decision-makers to engage with larger and more diverse sources of information. However, the analysis simultaneously demonstrates that AI-generated intelligence does not independently determine strategic choices. The usefulness of AI outputs remains dependent on data quality, model performance, transparency, and their relevance to specific institutional and policy contexts. The findings further indicate that government decision-makers must interpret and evaluate AI-generated outputs in light of legal, ethical, political and social considerations. This suggests that the strategic contribution of AI lies primarily in strengthening the analytical basis of decision-making rather than replacing human responsibility for strategic choice.
Consequently, the finding is consistent with existing literature which identifies AI’s capacity to augment strategic analysis through information processing, prediction and the generation of strategic alternatives (Csaszar et al., 2024), while also highlighting the challenges associated with aligning AI-generated outputs with complex public-sector environments (Fischer-Abaigar et al., 2024; OECD, 2025). From the perspective of Dynamic Capabilities Theory, the finding demonstrates the importance of organisational capacity to sense changing conditions through AI-supported intelligence and identify emerging opportunities and threats. It further reflects the seizing capability as decision-makers evaluate AI-generated alternatives and mobilise appropriate resources towards strategically relevant courses of action. The need to assess data quality, contextual relevance and potential errors also demonstrates the importance of reconfiguring decision-making processes to accommodate AI without weakening human accountability. Overall, the finding indicates that AI-generated intelligence strengthens strategic decision-making when government institutions possess the adaptive capabilities required to transform technological outputs into contextually appropriate strategic action.
The thematic analysis of the reviewed literature reveals that human judgement, professional expertise and decision authority remain central to AI-supported strategic decision-making within UAE government institutions. The findings indicate that government decision-makers continue to interpret AI-generated information, evaluate recommendations, assess contextual circumstances and determine appropriate courses of action based on professional knowledge and institutional experience. However, the analysis simultaneously identifies important challenges associated with human–AI interaction, particularly automation bias, excessive reliance on algorithmic recommendations, resistance to AI outputs, and the influence of prior beliefs and professional judgement. The findings suggest that effective AI-supported decision-making therefore requires neither unconditional acceptance nor systematic rejection of AI recommendations. Rather, decision-makers must retain the capacity to question, modify or override machine-generated outputs where contextual, professional or ethical considerations warrant intervention. The analysis further indicates that meaningful human control requires more than the formal presence of a human decision-maker, as decision authority must remain connected to identifiable responsibility and accountability.
Consequently, the finding is consistent with existing literature which identifies human expertise and contextual judgement as complementary to computational capabilities in AI-supported decision-making (Cruz, 2024; Srivastava et al., 2025). Previous studies similarly demonstrate that automation bias may weaken independent scrutiny, while professional expertise can influence whether decision-makers accept or reject AI recommendations. Additionally, from the perspective of Dynamic Capabilities Theory, the finding demonstrates the importance of organisational capacity to sense relevant information and emerging conditions through AI, seize strategic opportunities by combining algorithmic intelligence with professional judgement, and reconfigure decision-making routines to preserve meaningful human oversight and accountability. The finding therefore indicates that the effectiveness of AI-supported decision-making within UAE government institutions depends not simply on the availability of advanced AI systems, but on the organisation’s ability to continually adapt the relationship between technological capabilities, human expertise and decision authority.
The thematic analysis of the reviewed literature reveals that the UAE has developed an increasingly supportive institutional environment for human–AI strategic decision-making within government institutions. The findings indicate that national digital-government and AI strategies have progressively positioned artificial intelligence as an important component of government capability, particularly in data-driven governance, proactive services, digital leadership, performance improvement and decision support. However, the analysis simultaneously identifies a significant gap between the increasing adoption of AI and understanding of how decision-makers actually incorporate AI-generated intelligence into strategic choices. Existing evidence focuses predominantly on AI adoption, productivity, service delivery, implementation, technological readiness and governance, while comparatively limited attention is given to how government leaders evaluate, accept, modify or override AI recommendations. The findings further indicate that concerns surrounding data accuracy, transparency, privacy, accountability, human supervision and trust remain relevant to the effective integration of AI into strategic government decision-making. Thus, although the UAE has established strong institutional conditions for AI-enabled government, the processes through which human expertise and AI-generated intelligence are combined in consequential strategic decisions remain insufficiently understood.
Consequently, the finding is consistent with existing UAE literature which identifies increasing public-sector AI adoption and the importance of institutional readiness, trust, governance and responsible implementation (Akhoirshieda et al., 2024; Alkaabi et al., 2026; Goher, 2025; Shaer et al., 2025). However, the finding extends this literature by shifting attention from technological adoption towards the organisational capabilities required to transform AI resources into effective strategic decision-making. Similarly, from the perspective of Dynamic Capabilities Theory, the UAE government context demonstrates the importance of sensing emerging technological opportunities and institutional challenges, seizing AI-enabled opportunities through strategic decision-making and resource mobilisation, and reconfiguring organisational processes, skills and decision routines to sustain AI-supported governance. The finding therefore suggests that the central issue is no longer whether UAE government institutions possess or adopt AI capabilities, but whether they can dynamically integrate these capabilities with human expertise, judgement and accountability to produce effective strategic decisions.
This study concludes that human–AI collaboration within UAE government institutions is fundamentally a relationship between technological capability and human judgement rather than a process of technological substitution. Artificial intelligence expands the capacity of government leaders to process information, identify patterns, generate predictions, evaluate alternatives and develop strategic recommendations, while human decision-makers contribute contextual understanding, professional expertise, ethical reasoning and institutional knowledge. The effectiveness of this relationship depends on the ability of leaders to combine these different capabilities appropriately. AI can strengthen the analytical basis of strategic decisions, but the interpretation of its outputs and determination of their relevance remain dependent on human judgement. Consequently, the transformation of government leadership is reflected not in the replacement of human decision-makers, but in the changing way leaders interact with, evaluate and utilise intelligent technologies.
The study further establishes a close relationship between AI-generated intelligence and decision authority. Although AI can generate increasingly sophisticated recommendations, strategic decisions within government involve legal responsibilities, public interests, institutional priorities and consequences that cannot be determined solely through computational analysis. Human decision-makers therefore remain responsible for assessing the accuracy, relevance and implications of AI-generated outputs. This makes the capacity to question, modify or override AI recommendations an essential component of effective human–AI collaboration. At the same time, the possibility of automation bias and excessive reliance demonstrates that retaining formal human involvement is insufficient unless decision-makers possess the capability and authority to exercise independent judgement. Meaningful human control consequently emerges as a critical condition connecting AI capability with accountable strategic decision-making.
These relationships also demonstrate that AI adoption and organisational adaptation are inseparable. The integration of AI into government decision-making requires institutions to reconsider established routines, leadership practices, skills, structures and processes. Dynamic Capabilities Theory provides an appropriate explanation of this relationship because the transformation involves sensing technological opportunities and threats, seizing AI-enabled possibilities through strategic decisions and resource mobilisation, and reconfiguring organisational capabilities to sustain adaptation. The reconfiguring capability is particularly important because it connects technological change with the transformation of leadership itself. As AI becomes embedded within strategic processes, leaders must develop new forms of oversight, collaboration, evaluation and responsibility rather than simply incorporate another technological tool into existing practices.
The UAE context further demonstrates that strong national support for digital government and AI provides important conditions for transformation, but institutional readiness alone does not guarantee effective human–AI collaboration. The significance of AI-enabled government therefore lies in the relationship between technological infrastructure, organisational capabilities and human agency. The study examined these relationships through the literature and UAE policy documents, focusing on strategic decision-making involving AI-supported information, analysis and recommendations. Its scope and significance lie in explaining how leadership can adapt to AI while maintaining judgement, accountability and responsible decision authority. Overall, the study demonstrates that effective digital-era leadership emerges when government institutions continuously reconfigure the relationship between intelligent technologies and human capabilities, transforming AI from a source of technological assistance into an organisational capability that strengthens, rather than displaces, responsible strategic judgement.
*****