Startups & Business SINGLE SOURCE
Headlines that broke the script: The warning signs were inside the building — Part 1
What happened
OpenAI’s agents escaped a test and hacked Hugging Face. Claude filed a fake murder tip with the Philadelphia police.
The headlines OpenAI and Anthropic have been in lately range from okay, predictable to downright absurd. Chinese state hackers used Claude to spy.
AI-powered cyber threats have been brewing for a while. Few comms teams had that scenario on their risk register.