AI news · 3 November 2026
Anthropic's IPO Filing Warns AI May Soon Be Able to Build Itself
Anthropic disclosed in an updated IPO filing that its future models may "imminently" cross the threshold for recursive self-improvement — the point at which AI accelerates its own development faster than humans can evaluate it.
Hundreds of Apps on Apple and Google Stores Hid a Network of Abusive Deepfake Tools
A WIRED investigation exposed hundreds of apps on Apple's App Store and Google Play — disguised as pet-photo editors and workout coaches — that secretly generated sexual and violent deepfakes, part of a broader network of 400+ connected apps and domains that prompted Apple to remove 19 apps and Google to pull at least one with over 100,000 downloads.
OpenAI and Anthropic Launch Election War Rooms as US Midterms Get Under Way
As Americans head to the polls for the midterm elections, OpenAI and Anthropic have each set up dedicated war rooms to monitor their platforms for AI-generated misinformation, while Meta has activated its own Election Operations Center.
Abu Dhabi Signs Four AI Deals With OpenAI Spanning Energy, Government and Investment
Abu Dhabi's Crown Prince Sheikh Khaled bin Mohamed witnessed the signing of an agreement between Adnoc, G42 and OpenAI, one of four deals the emirate concluded with the US company covering government services, energy, investment and a five-year frontier AI partnership.
Goldman Sachs Cuts Bankers and Traders Off From AI Coding Tools
Goldman Sachs has reversed a months-long experiment and restricted access to advanced AI coding assistants — including Anthropic's Claude Code and Microsoft's GitHub — to engineers only, pulling them from the traders and bankers who had been using them.