AI Inc.: the drama, the money and the power plays
The fights, the exits, the filings and the price tags behind AI, and what they mean for the people on the other end. The facts come from the outlet linked on each story; Our read is our analysis, labelled as ours.
- The markdown
OpenAI halves the price of an AI it rates 'Critical' for hacking
OpenAI says GPT-6.1 Terra meets the top 'Critical' cybersecurity level in its own safety framework. Its API costs half what its predecessor did. Help Net Security
Our read: The same company sets the risk rating and the price. Cheaper access to a top-rated hacking capability widens who can use it, for defence and attack alike: the business logic and the safety logic point in opposite directions.
- The showdown
The FTC wants Anthropic's bosses under oath
The FTC sent 50-page demands to Anthropic and OpenAI for records on security incidents, safeguards, marketing and privacy, and wants testimony from Anthropic executives. It says more AI companies will follow. Axios
Our read: This landed the same week Anthropic filed to go public. Anything said under oath can end up in front of investors, so expect every lab's safety marketing to be rewritten by lawyers.
- The fine print
Anthropic tells investors an AI disaster could be too big to insure
Its IPO filing says no single insurer, "and potentially the insurance industry in the aggregate", may have the capacity to cover a catastrophic AI incident. The risk factors run past 90 pages. Implicator
Our read: When the seller says a risk can't be insured, the risk doesn't vanish. It lands on customers, shareholders and, in a big enough event, the public. Watch for insurers writing AI exclusions into business policies.
- The exit door
Microsoft offers buyouts to 60,000 people, a quarter of its staff
VP level and below in the US, Canada, India and the UK, per Bloomberg. Microsoft's memo calls it a new choice and doesn't mention AI. TechCrunch · TechCrunch
Our read: A buyout shrinks a company without the word 'layoff', and the people with the best options leave first. Apollo's chief economist finds no AI payoff yet outside the biggest tech firms: for most companies, AI is still a cost story.
- The hangover
$480 billion gone as China's AI stocks fall 40%
China's AI-heavy STAR50 index is down 40% since July. Goldman cut its target for Korea's Kospi by 25% as the AI trade unwinds; that index is down more than 35% from June. International Business Times AU · Investing.com
Our read: The AI boom's first big correction is hitting Asia, where retail buyers are heavily exposed. When chip-earnings expectations reset, the damage travels into savings and pensions, not just tech portfolios.
- The leak hunt
OpenAI checked fired safety researchers' phones, Journal reports
OpenAI security looked for Signal chats with Redwood Research's Ryan Greenblatt, per the Wall Street Journal. The researchers say they were never told what they leaked; Greenblatt calls the accusations false. The Next Web
Our read: Labs ask the public to trust their safety claims while treating safety staff who talk to outsiders as a leak risk. That tension is what regulators' document demands tend to surface.
Spot an error or something out of date? Tell us; corrections are made and logged. Written with AI assistance from public sources; checked against those sources before publishing. General information, not legal, medical or financial advice.