In April 2026, the United Arab Emirates announced a framework, under the direction of His Highness Sheikh Mohamed bin Zayed Al Nahyan, President of the UAE, and His Highness Sheikh Mohammed bin Rashid Al Maktoum, Vice President, Prime Minister and Ruler of Dubai, to deploy Agentic AI across 50 per cent of federal sectors, services and operations within two years – an initiative unprecedented in government.
Oversight rests with His Highness Sheikh Mansour bin Zayed Al Nahyan, Vice President, Deputy Prime Minister, while execution is coordinated by a taskforce chaired by His Excellency Mohammad Al Gergawi, Minister of Cabinet Affairs.
Weeks later, the Cabinet moved from ambition to architecture. It approved the governance framework defining the roles of every ministry and federal entity, the first list of services to transition under Phase One, and the largest training programme in the government’s history – equipping 80,000 federal employees with Agentic AI skills. Sheikh Mohammed declared that the journey to “UAE Government 4.0” had begun.
For the fire service, this is not a distant policy headline. Every federal entity, including the Civil Defence, must now form an implementation team headed by its leadership and begin transitioning services and operations to agentic models. In my previous article, I argued that fire leaders weighing AI investments should ask whether their strategy encompasses critical adoption phases. Now, the Cabinet has answered the question of national commitment. But more questions arise for the fire industry: Does the current infrastructure truly have the capacity for this transformation into AI agents? Do you have the required staff skills to support this journey? All of these are critical questions. The question that remains belongs to us: When the agent arrives, will your people know how to command it?
From digital to agentic: What actually changes
Agentic AI marks a categorical shift from the systems most fire services know today. Conventional digital tools, dashboards, alerts and records, inform a human who then decides and acts. An agentic system plans, decides, executes and improves towards a defined goal, escalating to a human only where judgement or authority is required. As Sheikh Mohammed put it, AI is no longer a tool – it analyses, decides, executes and improves in real time, acting as an executive partner to government.
From a scientific perspective, Russell and Norvig define an AI agent as an entity that perceives its environment through sensors and acts upon that environment through actuators to achieve specific objectives. Similarly, Wooldridge describes an intelligent agent as a computer system capable of autonomous action within an environment to meet its design goals. Unlike traditional AI systems that primarily generate predictions, recommendations or content in response to prompts, agentic AI can plan, execute multi step tasks, interact with external systems and evaluate outcomes with limited human intervention. In practical terms, an AI agent moves beyond informing decision makers to actively supporting the execution and management of operational processes within defined governance and oversight frameworks.
Despite increasing levels of autonomy, successful AI implementation remains fundamentally human centred. Humans define objectives, establish operational boundaries, validate outcomes and retain accountability for decisions. The most effective model is not human versus machine, but human and machine working together, combining the speed and analytical capability of AI with human judgement, experience and leadership.
The distinction is important. Traditional digital systems provide information for firefighters and officers to act upon; agentic systems can perceive conditions, evaluate options, initiate actions and continuously learn while operating within defined command, policy and governance boundaries.
Preparing for tomorrow
Consider a practical contrast. Today, a building monitoring dashboard alerts a duty officer that a water tank or fire pump in an industrial facility has developed a fault. The officer must then investigate the issue, assess the level of risk, and determine the appropriate course of action based on experience, operational judgement and established procedures.
Tomorrow, an AI agent detects the anomaly automatically, cross-references it against occupancy risk, inspection history and asset criticality, schedules the maintenance crew, issues the work order, verifies completion, updates the pre-incident plan and reports the closed-loop outcome to the officer. The human role shifts from processing information and coordinating actions to supervising outcomes and exercising command oversight.
The UAE is not starting from zero. Two decades of digital government, from early e-services to integrated identity systems and proactive service design, have built the data foundations agentic systems require. In the fire domain, the national smart alarm monitoring system Hassantuk already demonstrates the precursor logic: Sensors that detect, verify and dispatch with minimal human latency. Agentic AI extends that logic from a single life-safety function to the management of the fire service itself.
The research lens: Insights for accelerating agentic AI adoption
The Cabinet’s commitment to developing Agentic AI capabilities across 80,000 federal employees represents one of the most ambitious public sector workforce transformation initiatives undertaken globally. As organisations across government begin this journey, the findings of my doctoral research offer evidence-based insights that can help maximise the impact of these investments and support successful AI adoption.
The research found that facilitating conditions including infrastructure, technical support, data accessibility and leadership commitment play a critical role in enabling the successful adoption and integration of AI technologies. These findings suggest that training programmes achieve their greatest impact when delivered within an environment where personnel have access to the tools, resources and organisational support needed to apply new capabilities in practice.
The research also highlighted the importance of linking capability development with operational application. While formal training plays a vital role in building awareness, knowledge and confidence, the transition from learning to sustained use is strengthened when personnel are provided opportunities to apply AI tools within real operational environments. Officers who gain hands-on experience with AI-enabled systems are more likely to integrate these technologies into routine decision-making and become advocates for innovation within their organisations.
For fire service leaders implementing the national directive, the implication is clear: Training, technology, leadership and operational application should be viewed as complementary components of transformation. Pilot projects, operational demonstrations and structured exposure to AI-enabled workflows can reinforce learning outcomes, build trust and accelerate adoption. In this context, the national skills programme provides a critical foundation upon which operational transformation can be built, helping organisations move from understanding AI to confidently applying it in support of their mission.
Examples of the functions ready for agentic transformation
Where should a fire and rescue service begin? Before considering specific use cases, fire service leaders should recognise that successful agentic transformation is built upon digital maturity. Organisations that have already invested in digital platforms, integrated data environments, connected operational systems and data-driven decision-making are generally better positioned to realise value from AI investments. In many respects, agentic AI should be viewed not as a separate initiative, but as the next stage in the broader digital transformation journey.
While every organisation will have different priorities, several functions share common characteristics that make them well suited for agentic transformation: They are data-rich, process-driven, measurable and require continuous decision-making. These are the environments in which agentic systems can deliver the greatest value.
1. Plan review and permitting
AI-powered plan review platforms can assess building submissions against applicable fire and life safety codes, identify deviations, request additional information and route approvals to the appropriate authority. This can significantly reduce review times while allowing engineers and inspectors to focus on complex technical and judgement-based decisions.
2. Risk-based inspection targeting
Agentic systems can continuously assess risk by analysing factors such as occupancy changes, compliance history, inspection findings, sensor data, environmental conditions and community risk indicators. Inspection schedules can then be dynamically prioritised, ensuring that resources are directed towards the highest-risk occupancies.
3. Station readiness and resource management
Many fire services already use digital platforms to monitor staffing levels, apparatus availability, equipment status and operational readiness. An agentic layer can proactively identify gaps, initiate corrective actions, schedule maintenance activities, recommend resource substitutions and verify readiness before operational impacts occur.
4. Predictive dispatch and dynamic coverage
By analysing historical incident data, real-time risk indicators, weather conditions, traffic patterns and special events, agentic systems can model incident probability and recommend – or within approved parameters, execute – strategic resource repositioning. This allows operational coverage to adapt dynamically to changing risk conditions rather than relying solely on fixed response boundaries.
5. Firefighter health, safety, and wellbeing
Digital health and fitness monitoring systems increasingly track indicators such as physical fitness, medical assessments, fatigue and wellbeing metrics. Agentic systems can support workforce resilience by scheduling assessments, personalising fitness and wellness programmes, identifying emerging trends and alerting supervisors to potential risks before they affect operational performance.
6. Incident management and decision support
One of the most significant opportunities for agentic AI lies in incident management. By integrating operational resources, staffing information, apparatus status, building intelligence, geographic information systems, community risk data, weather conditions and live incident inputs, agentic systems can support commanders through real-time situational awareness, tactical recommendations, resource allocation, escalation planning and mutual-aid coordination. Human commanders retain authority and accountability, while agents provide continuous analysis and decision support throughout the incident lifecycle.
Agentic transformation at a glance
| Function | Today (digital) | Tomorrow (agentic) | Expected gain |
| Plan review | Engineers check drawings against code manually | Agent screens submissions, flags deviations, routes approvals | Faster permits, fewer backlogs |
| Inspection targeting | Fixed cycles and static priority lists | Dynamic risk-based scheduling generated continuously | Better risk reduction and resource utilisation |
| Station readiness | Platforms report gaps; officers initiate actions | Platform detects, initiates and verifies corrective action | Continuous readiness assurance |
| Dispatch and coverage | Static response zones and predetermined deployment models | Coverage repositions dynamically with modelled risk | Faster response times and improved coverage |
| Firefighter health and wellbeing | Periodic assessments and manual monitoring | Agent monitors trends, schedules interventions early | Improved wellbeing, fitness and injury prevention |
| Incident management and decision support | Commanders manually gather and assess information | Agent integrates data sources and provides real-time recommendations | Enhanced situational awareness and decision quality |
Critical considerations for fire leaders
While every fire and rescue organisation will progress at a different pace, several factors are likely to influence the success of agentic AI adoption. The findings from both practice and research suggest that fire leaders should consider the following:
- Establish clear governance and accountability. Agentic systems require defined authority levels, escalation pathways, audit mechanisms and human oversight. Leaders should ensure that responsibility for decisions remains clear, particularly in high-consequence operational environments.
- Define measurable operational outcomes. Technology initiatives should be linked to mission outcomes rather than technical capabilities. Whether the objective is reducing inspection backlogs, improving response times, enhancing firefighter safety or increasing operational readiness, success should be measured in operational terms.
- Start with targeted, low-risk use cases. Early implementation efforts should focus on functions where processes are well understood, data is available and outcomes can be measured. Demonstrating value in controlled environments helps build confidence and organisational trust.
- Invest in enabling conditions. Successful adoption depends on more than technology. Leadership support, data quality, infrastructure, interoperability, cybersecurity and access to technical expertise all play important roles in enabling operational use.
- Connect training with operational application. Capability development is most effective when personnel can apply newly acquired knowledge in real-world environments. Exposure to live systems, pilot projects and operational use cases helps transform learning into sustained practice.
- Build trust through transparency. Personnel are more likely to embrace agentic systems when they understand how recommendations are generated, what authority the system possesses, and when human intervention is required. Transparency supports confidence, accountability and responsible use.
- Maintain human command and oversight. Agentic systems should enhance decision-making, not replace professional judgement. Firefighters, officers and incident commanders remain responsible for operational decisions, while AI agents provide analysis, recommendations and process automation within defined boundaries.
- Scale based on evidence. Organisations should continuously evaluate outcomes, capture lessons learned and expand deployment where measurable benefits are demonstrated. Successful transformation is typically achieved through iterative improvement rather than large-scale implementation at once.
Lessons for the region
History suggests that transformative technologies rarely change organisations overnight. The steam engine transformed industry, electricity transformed productivity, and the internet transformed information exchange. Agentic AI has the potential to become the next general-purpose technology, fundamentally reshaping how organisations plan, decide and act. The challenge for fire service leaders is not whether this transformation will occur but how prepared their organisations will be to guide and govern it.
The UAE has volunteered to be the live laboratory of agentic government, and its fire and rescue services will be among the most scrutinised testbeds because nowhere are the stakes of autonomous decision making clearer than where seconds decide survival. Fire leaders across the Middle East should watch three things: How authority thresholds between agent and officer are drawn; how training is sequenced against live deployment; and how outcome metrics are reported. The lesson of the research holds at national scale: AI enhances missions, it does not replace grit. The agent will run the process. The human will command the incident. Departments that internalise that division of labour will find that the mandate was never a burden – it was a head start.
