Anthropic became the week's central character: a $965bn valuation that overtakes OpenAI, a $36bn debt raise, and a withheld cyber-defence model that the Bank of England, ECB, GCHQ and individual hackers all point to as a tool UK institutions cannot yet access. Around it sits the reckoning - softening jobs warnings, ROI doubts, and regulators moving to set the rules.Key stories:

• Anthropic's $965bn valuation overtakes OpenAI, backed by a $36bn Apollo–Blackstone debt deal

• Bank of England's Bailey warns UK banks are locked out of Anthropic's Mythos model

• GCHQ stands up agentic AI cyber defence as Iran turns Western models against the West

• ICO sets out a "safe AI innovation" plan as an EU law compliance study finds all 12 leading models failed

• Altman softens his jobs-apocalypse warning while Uber and Target question AI spend

• Fabricated AI citations hit a law firm, the IOPC and the courts in a single week

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit resultsense.substack.com