Recent Stories
The most consequential AI governance stories on our radar right now, with severity and jurisdiction. For the full day-by-day record, see the Daily Digest archive.
Last updated: 2026-07-15
China AI Companion Law Takes Effect: Doubao and Qwen Shut Down, Millions Lose Chat Data
China's Interim Measures on AI Anthropomorphic Interactive Services took legal effect today, forcing ByteDance's Doubao and Alibaba's Qwen to permanently close AI companion and persona features used by hundreds of millions, with Qwen permanently deleting user chat histories.
EU AI Act: Retroactive GPAI Fines Possible From August 2 as Only 10 Member States Show Enforcement Readiness
With 18 days to August 2, the EU Commission gains power to issue retroactive GPAI fines for violations dating back to August 2025, but only approximately 10 of 27 member states have built the national enforcement infrastructure needed to apply Article 50 transparency rules.
FTC Proposes Policy on AI Accuracy and Output Steering; Comment Deadline July 31
The FTC's proposed policy statement asserts that secretly steering AI outputs away from accuracy without disclosure may violate Section 5 of the FTC Act, with public comments due July 31 at regulations.gov (Matter No. P264200).
China AI Companion Rules Take Effect July 15; ByteDance Doubao Shuts Down Tomorrow
ByteDance will disable Doubao's AI companion and persona features on July 15 as China's Interim Measures on AI Anthropomorphic Interactive Services take legal effect, with Alibaba's Qwen having already removed humanlike interactive agents on July 10.
EU AI Act: 19 Days to August 2 GPAI and Article 50 Enforcement; Digital Omnibus OJ Pending
The EU AI Act's August 2 GPAI model obligations and Article 50 transparency enforcement date is now 19 days away, while the Digital Omnibus delaying high-risk AI obligations to December 2027 still awaits Official Journal publication.
ByteDance and Alibaba Shut Down AI Companion Features Ahead of July 15 Deadline
ByteDance is closing Doubao's AI persona creation feature on July 15 and Alibaba's Qwen and Tencent's Yuanbao have issued similar compliance notices, as China's Interim Measures on AI Anthropomorphic Interactive Services take effect - banning virtual companion services for minors and requiring CAC security assessments for platforms with over 100,000 monthly active users.
China MOFCOM AI Model Export Restrictions: No Final Decision After Six Days
Beijing's Ministry of Commerce consultations with Alibaba, ByteDance, and Z.ai on restricting overseas access to advanced Chinese AI models - including Qwen, Doubao, and GLM-5.2 - have produced no final policy decision as of July 13, with proposed measures ranging from simple filing requirements to domestic-only lockdowns for frontier models.
Apple Sues OpenAI for AI Hardware Trade Secret Theft
Apple filed a federal lawsuit in Northern California alleging OpenAI coordinated the theft of AI chip design and hardware trade secrets through systematic poaching of more than 400 engineers, with OpenAI's Chief Hardware Officer allegedly directing candidates to bring proprietary Apple parts to interviews.