AI at work
The Pentagon won a 2-1 ruling to ban Anthropic from defense work.
A US appeals court in Washington ruled on 25 September that the Pentagon may label Anthropic a "supply chain risk", the Associated Press reported.
Saturday, 26 September 2026
AI at work
Google released an AI face for companies that talks in 97 languages.
New York proposed fining OpenAI and rivals $25,000 per rule-breaking bot.
Musk said xAI now runs 550,000 Nvidia chips in one AI computer.
Google gave some Pixel 11 owners an AI that calls your business.
OpenAI is preparing a $500-a-month ChatGPT plan, BleepingComputer reported.
Meta's Zuckerberg rejected an AI slowdown. OpenAI's Altman and Anthropic's Amodei backed one.
OpenAI rival DeepSeek raised prices up to 4.5 times. Its CEO says demand held.
McDonald's said AI that tracks restaurant stock cut food waste by 15%.
Anthropic secured warrants covering up to 5% of cloud supplier Akamai.
Tesla builds hundreds of robots a week. Their AI still can't handle varied tasks.
Microsoft folded its consumer Copilot into one work app. Two years of effort ended.
Software business
Security
Video reports
The ERP LEADERS brief
The week in enterprise software, in one email
The stories that mattered, the programmes that failed and what they cost, and one case taken apart. Free to join, for anyone who works with ERP, SAP or enterprise AI. Read issue 01