THE PERMISSION [LAYER]

AI agent incident

GitBook says a targeted attack "using AI agents" reached its production infrastructure through an SSO flaw

Reported · 1 source

  • autonomous-attack
  • credential-theft
  • mcp

What happened

GitBook, a documentation platform, said an attacker exploited a flaw in its SSO login flow on November 2 to act as GitBook staff, reached internal code repositories through a GitHub App, obtained a Cloudflare API key and then full infrastructure access through a service account. GitBook disabled the affected service accounts, which took the product offline for about 16 hours until 11:56 UTC on November 3. Exposed data includes API tokens, integration tokens and secrets, and session signing secrets. GitBook says the activity focused on a single customer and describes the attack as highly targeted and driven largely by automation and AI; it has not published evidence of the tooling.

Why it matters

Documentation platforms hold tokens into source repositories and, through their MCP servers, into agents. GitBook tells customers to issue new API tokens, rotate third-party and integration secrets, and re-authenticate the GitBook MCP server.

New incidents by email

Get incidents like this one in your inbox

New AI agent incidents are covered in the briefing, with what to check on the agents you already run. At least once a week, free.

You confirm by email. Unsubscribe in one click.

Sources

  1. gitbook.com GitBook security incident, November 3, 2026

Background reading

What is an MCP gateway? An explainer on the problem behind this incident.

Covered in the briefing

Issue #4: A wave of AI platform breaches is after model keys November 10, 2026

Cite or correct this record

“GitBook says a targeted attack "using AI agents" reached its production infrastructure through an SSO flaw.” The Permission Layer AI Agent Incident Database, https://thepermissionlayer.com/incidents/gitbook-sso-breach-ai-agents/. Database updated November 10, 2026.

Every record traces to the dated sources above. If something here is wrong or missing, email [email protected] and it gets fixed. See this record in the full list.

The Permission Layer

Get the next incident like this explained

When an AI agent breaks out, leaks data or runs up a bill, the briefing covers what happened and what to check on the agents you already run. At least once a week, five-minute read, free.

You confirm by email. Unsubscribe in one click.