JUST IN: @Coinbase, @BitGo and 30+ others sign letter urging AI labs to give security researchers early access to frontier models, arguing attackers already do.

What happened

Coinbase, BitGo and more than 30 other companies signed a public letter asking AI labs to let security researchers access advanced AI models early, saying attackers already have that access.

Confirmed

Global impact / market context

Giving security researchers early look at powerful AI models lets them find flaws before attackers can use them, which could protect Coinbase, BitGo and other crypto platforms from fraud or theft, preserve user funds, and strengthen trust in the industry.

Analyst inference

Rapid advances in artificial‑intelligence models have sparked global worries about misuse, prompting regulators and tech firms to discuss safety rules, while the crypto sector already battles frequent cyber attacks, making the request for early security testing especially timely.

Analyst inference

What to watch

1. AI labs may adjust their model release policies to include controlled researcher access, which could create new collaboration frameworks and affect timelines for product launches. Proposed
2. Crypto firms might invest in internal AI security teams or partner with external experts, increasing spending on vulnerability research and potentially raising operational costs. Proposed
3. If AI labs grant early access, researchers could uncover defects that lead to patches, potentially lowering future breach risk for crypto platforms and improving investor confidence. Analyst inference

Evidence

• x.com