Abu Dhabi signs four OpenAI deals to make government machine-readable: keep a named person answerable

Abu Dhabi's Crown Prince, Sheikh Khaled bin Mohamed bin Zayed Al Nahyan, met OpenAI chief executive Sam Altman on 2 November and witnessed the signing of four agreements between Abu Dhabi entities and OpenAI, according to the state news agency WAM as carried by Gulf News. They cover government, energy and investment. The Abu Dhabi government and OpenAI will launch what they call a Machine-Readable Government initiative, which turns policies and regulations into a form AI systems can read so they can help officials assess compliance for permits and certificates. Officials keep oversight of final decisions. A first pilot is already running with the Abu Dhabi Agriculture and Food Safety Authority on food imports.

The other agreements put OpenAI's models into ADNOC's operational, investment and commercial decision-making, with G42's subsidiary Inception42; set up AI for investment analysis with Mubadala and Inception42; and commit Inception42 and OpenAI to a five-year agreement to support AI adoption across government, enterprise and education in the UAE. No financial terms have been published. Altman said Abu Dhabi and the UAE are setting a standard for AI integration across the public and private sectors.

This matters beyond Abu Dhabi because it is a government writing its own rulebook for a machine to apply. The people affected are not only civil servants. They are anyone who applies for a permit, a licence or a certificate, and anyone who works in the offices that decide them. The same pattern is arriving in schools, exam boards, councils and companies: rules get encoded, an AI applies them, and a person signs off. The risk is that the person signing off becomes a formality.

Dan's analysis is that encoding rules is a sound use of AI, and the pilot has the right shape: a narrow task, a named authority, and humans keeping the final call. The harder question is what happens to judgement. If staff stop doing the thinking, they lose the ability to spot when the machine is wrong. The education line in the five-year agreement carries the same test. Schools should ask what pupils and teachers will learn to do with AI, not only what it will do for them.

For school leaders, educators, parents and business leaders, the practical steps are these. Name one person who answers for each AI-assisted decision in your organisation. Write down which decisions a machine may recommend and which a person must make. Ask suppliers where data is processed and who can review a decision after the fact. And teach staff and pupils to check the output, because the check is the skill that keeps its value.

Dan Fitzpatrick is a former teacher and school leader who has spent years working with educators, schools and governments on adopting AI, and he writes on AI and the future of education as a Forbes contributor.

Dan Fitzpatrick is available for interview and comment on this story. Available today for live or pre-recorded radio, TV, podcast and print interviews, with a broadcast-quality remote setup for live or pre-recorded down-the-line interviews. Contact: dan@theaieducator.io