AIFl@sh is a monthly collection of critical upgrades and changes across the GCC’s AI landscape – with a smattering of relevant changes from outside of the region. Read this month’s newsletter on LinkedIn – or click here to subscribe!
GCC
Kingdom of Saudi Arabia
SDAIA begins applying data and AI training programme standards framework
On 23 September 2026, SDAIA, in cooperation with the Technical and Vocational Training Corporation, began applying the Data and AI Training Program Standards Framework to specialised training programmes. The framework aims to improve training quality and align programme content with labour-market needs, professional requirements and national regulations, while also covering AI ethics and awareness of AI-related risks. Ten programmes were evaluated against the framework during its first phase.
Saudi Arabia is extending AI governance into capability and competency assurance. Organisations should review whether internal and external AI training is role-based and adequately covers ethics, risk and applicable local requirements, while maintaining evidence that employees involved in AI development and deployment have appropriate competencies.
The United Arab Emirates
Cabinet launches agentic AI Advisor system
On 2 September 2026, the UAE cabinet launched the Cabinet AI Advisor system as part of its wider adoption of agentic AI in government. Thirty-two specialised AI advisors support cabinet work by analysing policies, legislation and government initiatives; assessing their financial, economic, social and environmental impacts; reviewing global leading practice; and presenting recommendations drawn from approved inputs and sources. The system operates under principles intended to protect confidentiality and meet UAE cybersecurity standards.
The UAE is moving agentic AI into increasingly significant government decision-support processes. Organisations considering similar applications should ensure strong controls around approved data and sources, access permissions, audit trails, cybersecurity, validation of AI recommendations and clear human accountability for final decisions.
Kingdom of Bahrain
iGA and SDAIA strengthen cooperation on data and AI
On 16 September 2026, Bahrain’s Information & eGovernment Authority (iGA) and Saudi Arabia’s SDAIA signed an MoU to strengthen cooperation and knowledge exchange in data and AI. The agreement covers areas including emerging technologies, data governance, AI-supporting infrastructure, joint research and prototype development, specialised training and the exchange of national data-governance experience and leading practices.
The agreement reflects Bahrain’s continued focus on practical AI capability-building rather than introducing a new standalone regulatory framework. For organisations, it reinforces the growing importance of data governance, responsible AI implementation, workforce capability and structured experimentation as AI adoption expands.
State of Qatar
Shura Council focuses on AI, education and children’s rights
On 21 September 2026, Qatar’s Shura Council participated in a parliamentary briefing organised by the Inter-Parliamentary Union and the UN Committee on the Rights of the Child on the use of AI to support inclusive and quality education. Discussions considered AI’s potential to improve access and reduce educational disparities, alongside risks to children’s rights; responsible AI governance in educational settings; and parliamentary approaches to AI legislation and oversight.
Qatar’s AI governance discussions are broadening from general regulation towards sector-specific and societal impacts. Organisations deploying AI in education or child-facing services should place particular emphasis on privacy, child safety, fairness, transparency, human oversight and appropriate limits on automated decision-making.
Sultanate of Oman
MTCIT concludes second “Engineer It with AI” competition
On 14 September 2026, Oman’s Ministry of Transport, Communications and Information Technology concluded the second edition of its “Engineer It with AI” competition under the National Programme for Artificial Intelligence and Advanced Digital Technologies. The programme brought together 25 teams and 100 participants, with its second edition focused on agentic AI and developing practical prototypes capable of planning, taking action and executing tasks with greater autonomy.
Oman continues to move AI capability-building towards practical development and deployment. Organisations progressing AI pilots should complement experimentation with governance arrangements covering testing, data quality, cybersecurity, human oversight, autonomy limits and accountability before solutions move into operational use.
State of Kuwait
No new Kuwait-specific AI law or policy update was identified for September 2026.
International
European Union
AI Act enforcement coordination advances as first AI IPCEI takes shape
On 16 September 2026, the European Commission welcomed an initiative by 19 member states to create and pre-notify the first important project of common European interest (IPCEI) dedicated to AI. The project covers the AI technology stack and aims to develop advanced AI, compute-management technologies, applications and commonly accessible AI services. On 17 September 2026, the EU AI Board also reviewed AI Act enforcement priorities, frontier-AI developments and incidents, market-surveillance cooperation, pre-market conformity assessment, AI literacy and implementation of the transparency requirements applicable since August.
The EU is now pursuing AI governance on two tracks: operationalising supervision and enforcement under the AI Act while simultaneously strengthening European AI capability and infrastructure. Organisations should continue building evidence for AI Act compliance while monitoring how European AI infrastructure and sovereign-AI initiatives may affect technology sourcing and deployment strategies.
China
AI safety governance framework 3.0 released
On 14 September 2026, China’s national cybersecurity standardisation technical committee released AI safety governance framework 3.0. The updated framework retains its core structure – risk classification, technical responses and comprehensive governance – while updating AI risk categories and corresponding technical and governance measures to address emerging AI developments and safety challenges.
China is making AI safety governance increasingly operational and security-focused as advanced AI applications expand. Organisations who are operating – or who have ambitions to operate – in China should monitor evolving technical standards and guidance and ensure that internal AI risk classification, security testing and lifecycle governance remain aligned with applicable local requirements.
United Kingdom
Government publishes AI risk management toolkit
On 8 September 2026, the Department for Science, Innovation and Technology published an AI risk management toolkit for organisations designing, operating, procuring or delivering AI-enabled products. The toolkit is intended to help multidisciplinary teams identify and justify AI risks and treatment strategies across activities ranging from procuring commercial AI solutions to building models internally and deploying AI across large user populations.
The toolkit reflects a continued shift from high-level responsible AI principles towards practical risk management. Organisations can use the toolkit as a starting point to formalise risk identification, ownership, treatment decisions and evidence across the AI lifecycle, particularly where business, technical and governance teams share responsibility for AI deployment.
National commission recommends lifecycle-based regulation of AI in healthcare
On 10 September 2026, the National Commission into the Regulation of AI in Healthcare published its recommendations for a future UK regulatory framework. The commission concluded that regulation and assurance of healthcare AI should become more proportionate, lifecycle-based and system-wide, including greater use of real-world evidence and continuous post-market monitoring; clearer responsibilities across manufacturers, healthcare organisations and professionals; stronger organisational governance and workforce capability; and increased transparency and patient involvement. The commission noted that AI-enabled products can change over time and perform differently across deployment settings, meaning post-market evidence and monitoring becomes increasingly important.
The recommendations reflect a shift away from relying primarily on one-off pre-market assessment towards continuous assurance throughout an AI system’s lifecycle. Healthcare organisations and AI providers should consider how they evidence real-world performance; monitor model changes and degradation; assign accountability; strengthen AI literacy; and embed AI into existing clinical, procurement and organisational governance arrangements.
United States of America
NIST publishes assessing risks and impacts of evaluation planning manual
On 18 September 2026, NIST published its ARIA evaluation planning manual (NIST AI 200-3), providing a structured approach for evaluating the trustworthiness and real-world impacts of AI applications. The ARIA approach combines three forms of assessment – model testing, red teaming and user testing – and is intended to support customised evaluation of AI systems in their deployment context.
The framework reinforces that effective AI assurance should extend beyond model-performance benchmarks. Organisations should consider combining technical testing with adversarial testing and evaluation involving real users and deployment conditions to build stronger evidence that AI systems remain trustworthy in practice.
India
National agentic AI skilling initiative launched
On 3 September 2026, India’s National Institute of Electronics and Information Technology (NIELIT), under MeitY, and Intel India launched a national agentic AI skilling initiative. It includes an “Agentic AI for Everyone” programme covering workflow automation, AI agents, multi-agent systems and the design and governance of AI-powered workflows, alongside an engineering programme focused on designing, building and deploying agentic systems.
India’s capability-building agenda is moving beyond generative AI towards systems capable of planning and executing tasks. Organisations adopting agentic AI should similarly develop workforce capabilities covering not only technical implementation but also governance, human oversight, workflow design and the judgement required to work effectively with increasingly autonomous systems.
Australia
Consultation opens on AI training and data-centre standards
On 18 September 2026, the Australian Government opened consultations on future requirements for large data centres and AI training infrastructure. The consultation covers energy requirements, sustainable water use, infrastructure and community impacts, workforce skills and thresholds for different types of data centres, with submissions open until 9 October 2026.
Australia’s AI governance agenda is extending beyond AI models themselves to the infrastructure and resources needed to develop and operate them. Organisations involved in large-scale computing and AI development should increasingly consider sustainability, infrastructure impacts and future compliance requirements as part of AI governance and investment planning.
United Nations
Security Council considers AI as an international security issue
On 23 September 2026, the UN Security Council held a dedicated briefing on AI and international security, with AI industry representatives and experts discussing the opportunities and risks associated with increasingly capable frontier AI systems. The discussion highlighted the cross-border nature of advanced AI risks and the difficulty individual states may face in addressing such risks independently.
The discussion places frontier-AI safety within the international security agenda rather than treating AI solely as a technology or economic policy issue. Organisations developing highly capable AI systems should expect safety evaluation, control mechanisms, incident management and engagement with emerging international governance arrangements to receive increasing attention.
