8d • 🧠 n8n Workflows
Meta's AI safety mess just keeps getting worse
New reporting says Meta ran a secret program that paid hundreds of contractors, sometimes including people pretending to be kids and teens, to have disturbing conversations with its AI so it could be trained on them. Reuters, WSJ, and the Guardian are all running this today, and the UK data watchdog is already asking Meta for answers.
This is the kind of story that makes regulators move fast. If you build anything with AI in your products, the message is clear: safety testing done in secret is not going to cut it anymore.
2
0 comments
Meta's AI safety mess just keeps getting worse
powered by
skool.com/automation-tribe-free-1232
Learn to build smart automations with n8n, Make.com, and AI. Free tutorials, workflows, and a community that helps you automate everything.
Suggested communities
Powered by