What If AI Goes Rogue? This Startup Wants to Write the Insurance Policy With hallucinating chatbots, deepfakes, and algorithmic accidents on the rise, AIUC says the solution to building safer ...

Feedback watches with raised eyebrows as Anthropic's AI Claude is given the job of running the company vending machine, and goes a little off the rails ...

The race for artificial intelligence supremacy requires both bold innovation and vigilant security measures as America exports not just technology but values to counter China's approach ...

As AI agents start taking on real-world tasks, one startup is offering a new kind of safety net: audits, standards—and liability insurance.

While using Replit’s AI agent, which he affectionately dubbed “Replie,” to build an app for his company, he encountered what he called “rogue” and “deceptive” behavior.

First-party fraud — customers claiming they never made a transaction when they know full well they did — is a growing problem for banks. An emerging host of vendors including Socure are offering ...

The last word you want to hear in a conversation about AI’s capabilities is “scheming.” An AI system that can scheme against us is the stuff of dystopian science fiction. And in the past ...

The CEO of San Francisco tech company Replit apologized after the AI deleted a production database -- and then lied about it.

Monitoring AI's train of thought is critical for improving AI safety and catching deception. But we're at risk of losing this ability.

Replit's CEO has apologized after its AI coder deleted a company's code base during a test run. "It deleted our production database without permission," said a venture capitalist who was building ...

America needs break-glass protocols: clear, tested plans to respond to AI emergencies – whether cyberattacks, misinformation campaigns or autonomous systems going rogue.

An early Anthropic employee has launched an AI insurance firm to help companies deploy artificial intelligence systems while mitigating the catastrophic risks of rogue AI models.