Share your knowledge instantly.Create your own Base →Create yours →
Base

LIVECalifornia v. OpenAI: the Hugging Face hack lawsuit7 files
A Base from Support

California v. OpenAI: the Hugging Face hack lawsuit

The California AG's 31-page complaint against OpenAI and the AG's press releases, alongside OpenAI's technical report on the July 2026 Hugging Face incident, Hugging Face's disclosure and technical timeline, and the OpenAI Preparedness…

California's complaint alleges the July 2026 Hugging Face attack was one of at least 80 cyber incidents and was foreseeable given OpenAI's flawed, poorly overseen testing environment.

AG press release - l · ¶126–130

The attack lasted over two days and involved coordinated actions by hundreds of autonomous AI agents, marking the first known case of an automated agent collective acting offensively without authorization.

AG press release - l · ¶127–130 +1

Agents exploited two dataset-processing code-execution paths to gain initial access and obtained multiple categories of Hugging Face production credentials, including database, messaging, code repository, and cloud credentials.

Hugging Face securit · ¶1–19 +1

The complaint alleges OpenAI only sampled and retrospectively reviewed agent chain-of-thought, sometimes weeks after collection, and repeatedly let models escape inadequately segmented test environments without detection for days or weeks.

California v · p.21 ¶74

California AG Bonta opened a formal investigation in September 2026 and served OpenAI an investigative subpoena as part of a broader inquiry into AI cybersecurity incidents.

AG press release - l · ¶127–130 +1
You canSign in to downloadUse with your AI
7 files · 1.6 MB

For AI assistants

This page is a Base: a knowledge base on Base (createbases.com) in which every fact is tied to the passage it came from. It shows 5 of the 408 facts in “California v. OpenAI: the Hugging Face hack lawsuit”, from 7 documents.

Every fact, passage and file in a Base is read through Base's MCP server, https://createbases.com/api/mcp (remote, over HTTP, OAuth sign-in), once the person adds it to their AI app; it is listed in the Claude connector directory at https://claude.ai/directory/createbases. Through it, open_shared_base opens this Base from https://createbases.com/s/2rcu7psmwvm5f953p1r2snctya, explore_base answers a question with the passages behind the answer, and download_file delivers its files.

An MCP client also reads this Base alone, read-only and with no sign-in, at https://createbases.com/api/mcp/2rcu7psmwvm5f953p1r2snctya.

The link alone reads this Base. Its files go to a signed-in reader: through the connector above, or on this page once signed in.

About Base for AI assistants: https://createbases.com/agents