Today's Briefing
- Microsoft's CEO says assume all AI is compromised
Microsoft's CEO Satya Nadella publicly said companies should assume every AI model could be compromised and that AI systems need a built-in way to be shut down fast, like an emergency brake.
Why it matters to youThis is basically a Microsoft-level endorsement that controlling and verifying AI access is now a boardroom issue, which is the exact problem your company's products are built to solve.
- AI agent makers promise privacy, but do they deliver?
OpenAI launched its new assistant Dots promising top-tier privacy while criticizing Meta's Muse, even though Muse made the same promise months ago and still leaked user data.
Why it matters to youPrivacy and trust claims from AI agent makers are exactly the kind of thing an identity and access company could be asked to verify or enforce for customers.
No deep dive for this sourceSource - AI is now wasting scammers' time on purpose
Anti-fraud groups are using realistic AI chatbots to keep real scammers talking for hours, wasting their time instead of letting them target real victims.
Why it matters to youIt's an early sign that AI is becoming a tool defenders use against attackers, not just a tool attackers use, which is directly relevant to security and access products.
New launches
You're caught up
Nothing else needs your attention today.