Associated Press reported that China’s rapid public and commercial adoption of AI is shaping how the technology may be used globally. The report links mass use, state support and platform integration to a broader governance question about how national AI ecosystems influence global norms. 

Associated Press reported that Apple agreed to a 250 million dollar settlement in a class action over alleged false advertising of Siri’s AI capabilities. The claim centred on Apple Intelligence features promoted with the iPhone 16, making the dispute relevant to AI product claims and consumer-facing reliability. 

Reuters reported that the US government expanded AI stress testing to include unreleased models from Google DeepMind, xAI and Microsoft. The testing is led by CAISI and focuses on cyber, chemical and biological risk, data integrity and pre-release model vulnerabilities. 

Regulation

  • European Parliament IMCO listed a 6 May discussion on safety and cybersecurity risks potentially posed by advanced AI systems, including Anthropic’s Mythos. The order of business frames the discussion through the EU AI Act, the Cyber Security Act and other European digital rules. 

Cases

  • CourtListener reflects the latest activity in Andersen v Stability AI Ltd, including DeviantArt materials concerning sealing around Surovenko’s participation in the development of DreamUp. The connected discovery record shows a live dispute over Midjourney source code for its ‘overfit’ filter and ‘style reference’ materials, with plaintiffs presenting the code as relevant to copying, fair use and style mimicry while Midjourney resists further production as unnecessary and commercially sensitive. 

Events

  • UNIDIR will hold the Global Conference on AI, Security and Ethics 2026 in Geneva and online on 18 and 19 June 2026. The conference will examine AI’s implications for international peace, security, ethics and responsible governance. 

Sources: Associated Press, Reuters, European Parliament IMCO, CourtListener, Springer, UNIDIR

Tags
Siri
mythos
DreamUp
Overfit
Oversight
cybersecurity