Category: AI

Salesforce's Marc Benioff confirmed his company will spend $300 million with Anthropic in 2026 on AI coding agents. The cheque is the signal: every major SaaS vendor now sees AI as where the next decade of value will be created and captured. If you run a serious SAP estate, much will turn on whether you can see what AI agents do inside your business, and govern what they are allowed to do.

The CEO of SAP, addressing half a million LinkedIn followers earlier this year, was making the case for the Autonomous Enterprise. That's a world where AI agents run business processes end-to-end. Closing periods. Chasing approvals. Reconciling accounts. Acting, in his words, autonomously.

This week, Nvidia bought an open-model library.  AT&T revealed it has walked away from expensive frontier AI for cheaper alternatives.

Capita missed two recovery deadlines and pensioners are waiting months for bereavement payments. The reason: 20 million missing records. Everyone's got a plan until they meet the data.

Scott Bessent called an emergency meeting with Bank of America, Citi, and Wells Fargo last month. Anthropic's Mythos can find critical software vulnerabilities at a speed and scale no human team can match. For organisations running SAP, this is a cybersecurity threat that cannot be ignored.

Palantir's CEO claimed AI agents could compress SAP migrations from years to weeks. Markets lost $300 billion in a single session. But now 38 researchers from MIT, Harvard, and Stanford have shown AI agents can lie to their handlers. The findings will give every SAP leader pause.

OpenAI is worth $500 billion. SAP has 50 years of enterprise data. Guess which survives?