New York Magazine has named 26 people it considers unusually influential in AI, from frontier-model safety leads to chip analysts and political figures. The list’s useful point is that power in AI does not always follow the company org chart, or the fame chart.
Watch Desk analysis
What happened
The magazine assembled its 2026 AI Power List with Transformer, drawing on its reporting and new interviews with more than 50 senior industry figures. It spans private and public sectors, and includes founders, engineers, lawmakers and regulators.
The profiles in the supplied article show different routes to influence. Anthropic’s Logan Graham leads its Frontier Red Team, which tests models for dangerous capabilities. Paul Christiano helped develop reinforcement learning from human feedback at OpenAI, later founded the Alignment Research Center and joined OpenAI’s nonprofit board. SemiAnalysis founder Dylan Patel is profiled for research that connects chip and data-centre details to business consequences. Dwarkesh Patel’s influence, the magazine argues, comes from a podcast listened to by Silicon Valley leaders.
The article describes a list that is overwhelmingly male, with an average age of 36. It also notes that people who shape AI may sit outside the most obvious positions of authority: a specialist, researcher or interviewer can sometimes have more sway than a senior executive.
Why it matters
AI decisions are shaped not only by who owns the largest companies, but by who develops techniques, sets safety priorities, interprets the hardware bottlenecks and influences public or political debate. Knowing those routes to influence gives readers a better map of the industry than a roll-call of chief executives alone.
The list is an editorial selection, not a measured ranking of power. Its value is in the profiles and the arguments behind them, which invite scrutiny of who gets heard and who is missing.
Our read
The most interesting question is not whether every name deserves a place. It is what the list counts as influence: technical contribution, control over deployment, public persuasion, or access to decision-makers. Those forces overlap, but they are not interchangeable. A person who can change how a model is tested may matter as much as one who can approve its launch, even if fewer people recognise the name.
What to watch
- Whether the profiles’ claims about influence are borne out by later decisions and technical developments.
- Which less visible researchers, advocates and public officials gain influence as AI policy and deployment mature.
- Whether future lists give a fuller account of whose work and perspectives remain under-represented.
Discussion spark: Should influence in AI be judged mainly by who controls companies and capital, or by who shapes the technology, its safeguards and the public debate?
Sources and evidence
- Source update (2 November 2026, 10:00 UTC)
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.