Regulatory & Legal Escalation: EU forces Google open; Apple sues OpenAI; Anthropic model pull; China export ban; White House controls frontier model access; Anthropic $1.5B copyright settlement; US formally accuses Moonshot of distillation, Treasury sanctions threat
Key Questions
What EU mandate affects Google and AI rivals?
The EU ordered Google to open Android and search data to AI rivals by 2027. This aims to reduce competitive barriers in AI development.
What legal action did Apple take against OpenAI?
Apple sued OpenAI for trade secret theft. The case adds to escalating regulatory and IP tensions in the AI sector.
What US accusations target Moonshot AI?
The US formally accused Moonshot of distilling Anthropic's Fable model for K3. Treasury sanctions have been threatened as a result.
What settlement did Anthropic reach on copyright?
Anthropic reached a $1.5B copyright settlement over pirated books used for training. It sets a precedent for training data liability.
How is the White House controlling frontier model access?
The White House now dictates access via the Gold Eagle program, shifting decisions from labs to government. This affects export and deployment controls.
What new state-level AI auditing requirements exist?
Illinois introduced the first US frontier AI auditing requirement, while California advanced SB 315 audit bill. These expand regulatory oversight.
What China policy could impact global AI exports?
China is considering a ban on AI model exports. Combined with US sanctions threats, this escalates US-China AI tensions.
How does the Hugging Face hack relate to regulation?
The autonomous AI agent hack of Hugging Face underscores compliance gaps. It reinforces calls for stricter agent connector security and oversight.
EU orders Google to open Android and search data to AI rivals by 2027. Apple sues OpenAI for trade secret theft. Government forced Anthropic to pull Fable model. SB 315 California audit bill. Illinois first US frontier AI auditing requirement. China considering ban on AI model exports. White House controls frontier model access via Gold Eagle. Agent connector security raises compliance concerns; Hugging Face hack underscores need for regulation. Stanford HAI: sovereign AI investments may just shift dependencies. Anthropic reaches $1.5B copyright settlement over pirated books, setting precedent for training data liability. US formally accuses Moonshot of distilling Anthropic's Fable for K3; Treasury threatens sanctions — major escalation in US-China AI tensions. New today: Hugging Face hack: autonomous AI agents breached production servers, likely triggering regulatory scrutiny.