Breaking news, AGI updates, and analysis of the intelligence revolution.
📰 News South Africa's draft National AI Policy was withdrawn after someone noticed its academic citations were entirely fabricated by AI. The policy was, in part, about the dangers of AI-generated misinformation. You cannot make this up.
📰 News OpenAI starts a $4B company to help businesses deploy AI. The EU just postponed high-risk AI rules for over a year. Colorado's hard-fought AI regulation was replaced with a watered-down version. The US government will pre-test frontier AI models before release. And Google caught the first AI-developed zero-day exploit.
📰 News EY's 44-page loyalty fraud report had 60% fake citations, a non-existent McKinsey study, and broken links to Forbes and WIRED articles that never existed.
📰 News OpenAI's Daybreak AI cybersecurity platform arrives as Google confirms it stopped the first known AI-developed zero-day exploit. Two stories, one trend: AI security is now both weapon and shield.
📰 News Anthropic writes two $1.5B and $200M cheques in the same week, UK government launches a Claude-powered chatbot, OpenAI and Apple head for a messy legal breakup, Google confirms the AI cyber offensive has begun while OpenAI launches a defensive platform the same day, the EU softens its own AI Act, and China writes the world's most detailed agentic AI rules.
📰 News NZDF deployed Copilot at 'speed' across every device in September. Eight months on, there's still no AI directive. Staff can upload RESTRICTED documents. The risk assessment warned about it a year ago.
📰 News OpenAI faces a class-action lawsuit claiming ChatGPT sent user queries and personal data to Meta and Google through embedded tracking pixels — turning intimate AI conversations into ad-targeting fuel.
📰 News Trump and Xi face off over AI in Beijing, OpenAI stuns by proposing a global watchdog with Chinese participation, the UK reforms its cyber laws for the AI era, Microsoft's 100-agent security swarm beats Anthropic's Mythos, Anthropic takes the enterprise crown from OpenAI, and Meta promises the first truly private AI chat.
📰 News Ontario approved 20 AI medical scribes for doctors. The auditor found 60% recorded wrong drugs, 9 fabricated treatments, and accuracy scored just 4% in procurement.
📰 News The highest-profile AI partnership in consumer tech is fracturing. OpenAI says Apple buried ChatGPT behind friction, failed to promote the integration, and left a deal worth billions producing pennies.
📰 News APAC governments are racing to build sovereign AI systems, but NZ's new AI Blueprint sidesteps the infrastructure question. The gap between neighbours and Aotearoa is widening.
📰 News A former NASA scientist just launched a startup claiming to generate electricity from the quantum vacuum itself — with $12M in funding, an NSF grant, and a peer-reviewed paper backing the physics.