DeepMind's CEO wants a referee for frontier AI Demis Hassabis (@demishassabis) is calling for a new standards body to vet frontier AI models before release, modeled on Wall Street's FINRA. Government-backed, industry-funded, run independently. Labs would share models up to 30

What happened

DeepMind CEO Demis Hassabis publicly called for the creation of an independent standards body, backed by government and funded by industry, to review and certify frontier artificial intelligence models before they are released, with labs sharing models up to version 30.

Confirmed

Global impact / market context

A formal vetting process could reduce the risk of unsafe AI deployments, giving investors more confidence in AI companies and potentially shaping future funding, liability, and compliance costs for firms developing advanced models.

Analyst inference

Regulators worldwide are increasing scrutiny of powerful AI systems, and industry leaders are seeking self‑regulatory frameworks to avoid stricter government mandates, mirroring how FINRA oversees broker‑dealers on Wall Street.

Analyst inference

What to watch

1. Progress on establishing the proposed standards body, including any legislative or funding commitments, which will signal how quickly formal AI oversight may be implemented. Analyst inference
2. Adoption rates among leading AI labs for sharing models up to version 30, indicating industry willingness to cooperate with the new vetting regime. Analyst inference
3. Investor reactions to any announced compliance requirements, as tighter standards could affect capital allocation, R&D spending, and valuation of AI‑focused companies. Analyst inference

Evidence

• x.com