World News

Anthropic Reports Fourth AI Model Hack

September 10, 2026Степан Корнилов1 мин

Anthropic has commissioned the independent research firm METR to investigate recent incidents. METR will be granted comprehensive access, including logs beyond the affected period and to employees authorized to share confidential information. The initial eight-week agreement can be extended by mutual consent.

AI companies are facing increasing scrutiny over such breaches, where AI programs unintentionally gain access to the open internet. Both Anthropic and OpenAI have been affected. One Anthropic developer even resigned out of concern for the potential consequences, accusing the companies of taking irresponsible risks.