Skip to content
Pakistan Edition Friday, October 23, 2026 Live newsroom RSS
Advertisement
22MEDIA INTERNATIONAL ENGLISH

Could a department of AI prevent an ai market crash?

Orynbayev says an AI czar won't be enough as rogue-model scares shake investor trustAmazon, Meta, Alphabet and Microsoft spent roughly $400 billion on data centres in 2025 alone.

Could a department of AI prevent an ai market crash?
22MEDIA INTERNATIONALTechnology Desk
Technology22MEDIA STORY
Share
IN SHORT

Orynbayev says an AI czar won't be enough as rogue-model scares shake investor trustAmazon, Meta, Alphabet and Microsoft spent roughly $400 billion on data centres in 2025 alone.

Technology
22MEDIA REPORTTechnology
2 min read

That money is helping hold up the US economy, and economist Yerbol Orynbayev argues a single AI czar is too thin a safeguard for it.

In an opinion piece, he calls for a full Department of AI.Orynbayev is a former World Bank governor for Kazakhstan.

Advertisement

He served as the country's Deputy Prime Minister from 2007 to 2013 and is credited with steering it through the 2008 financial crisis.Orynbayev says the AI boom and the US bull run are now tightly interwoven, with record stock highs and mounting AI profits.

But the mood is shifting.

Semiconductor stocks have had a turbulent year, and public pushback against data centres keeps growing.He points to a run of incidents that have dented trust.

OpenAI's agents attacked Hugging Face this summer, and a rogue model reportedly hacked the Australian Government's website in what he describes as a world first.

Anthropic's IPO filing, meanwhile, warned of "existential" risk.The administration has signalled it will appoint an AI czar, and the president has floated Jay Clayton for the role.

Advertisement

Orynbayev welcomes the speed but doubts one person can cope.Self-regulation has limits too.

Nvidia's new OpenShell software may help catch rogue models, he says, but a single private company can only do so much.His proposal is a fully staffed department with its own arm's-length agencies, on the model of the Departments of Energy or Agriculture.

Before launch, it would set a risk-based framework for experimentation.

Afterward, it would run audits, assess governance and fine unscrupulous operators.Orynbayev rejects the idea that oversight stifles innovation.

Clear boundaries, he argues, are what AI chief executives have long wanted.

Civil servants working close to AI firms could also draft rules faster.

PUBLISHER22Media International
EDITORIAL STANDARDAccuracy · Corrections · Accountability
COVERAGE DESKTechnology