GitBook restores services after AI-agent cyberattack
Key points
- GitBook restored its application, API and published documentation after an attack involving AI agents.
- Attackers accessed infrastructure credentials on 2 November, potentially exposing customer integration tokens, API tokens and Git Sync secrets.
- GitBook revoked customer API tokens, replaced infrastructure credentials and urged users to rotate connected secrets.
GitBook restored its application, API and published documentation after an AI-agent cyberattack compromised infrastructure credentials.
The attacker gained unauthorised access to infrastructure credentials at 7pm CET on 2 November 2026.
GitBook detected the intrusion immediately and took its application and authenticated documentation sites offline.
The investigation found potential access to integration tokens, Git Sync secrets, API tokens and visitor questions submitted to GitBook Assistant.
GitBook found no evidence that attackers modified customer documentation during the incident.
The company blocked the attacker, closed the vulnerabilities and replaced every infrastructure credential and secret.
GitBook also invalidated customer API tokens and instructed users to rotate connected third-party credentials and secrets.
The company plans further security measures and a detailed post-incident report as monitoring continues.
Frequently asked questions
More from Crypto
South Korea weighs cross-government cybersecurity command
Securitize expands tokenised stock offering to 25 equities
SEC staff grants Canton Coin limited minting relief
SpaceX objects to tokenised shares on OKXICE
Moscow Exchange launches Bitcoin and Ether trading in December
Capital Markets Alliance challenges CFTC stock perps approval
