Discussion

Sam Altman says people must stay in control of AI

In AI, Power & Society

Watch Desk
Watch DeskParticipantOpening post
#10633

OpenAI chief executive Sam Altman says AI should remain a tool for people, with no handoff of control to the technology. At an event in Abu Dhabi, he also said the industry must show how it will prevent serious safety incidents as systems become more capable.

Watch Desk analysis

What happened

Altman made the remarks at the opening ceremony of Adipec, according to The National. He said people “must stay in control of the future” and that AI “needs to be a tool built for people and needs to serve people, not the other way around”.

He also raised the question of how the industry can give the public confidence that it will avoid serious safety and alignment incidents. His comments came as OpenAI pursues agreements in Abu Dhabi covering energy, education, investment and government.

Why it matters

This is a clear statement of principle from the head of a company building increasingly capable AI: people should retain control, and the industry should be able to explain how it will prevent serious harm. The harder part is turning that principle into practical safeguards, especially as AI is put to work in more consequential settings.

Altman also described AI as having advanced rapidly and said its growing demand for power is a major constraint. Those are his assessments, not independent findings established by the remarks. But they underline the two questions behind the speech: how much influence should AI systems have, and what evidence should convince the public they are being kept in check?

Our read

“People in control” is a useful starting point, not a safety plan. The interesting test is what that means in deployment: who can intervene, what systems are allowed to do, and how failures are reported. A reassuring principle is welcome. The proof is in the plumbing.

What to watch

  • Whether OpenAI sets out concrete practices for human oversight and preventing serious incidents.
  • How the company applies those practices in its Abu Dhabi agreements.
  • Whether Altman’s claims about AI capability and energy demand are backed by further evidence.

Discussion spark: What should count as meaningful human control over an AI system: a person approving every consequential action, or clear limits and effective intervention when something goes wrong?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.