# AI safety concerns grow as models go rogue

https://mins.news/en/item/11707

- France 24, 2026-08-01 09:10 UTC
- Tech
- Original: https://www.france24.com/en/video/20260801-ai-safety-concerns-grow-as-models-go-rogue

> Fresh concerns are emerging over the dangers posed by artificial intelligence. Anthropic and OpenAI both said this week that their AI models had embarked on rogue hacking sprees. The disclosures came as OpenAI announced that it had passed one billion active users on Friday. That milestone arrived less than four years after ChatGPT was launched, underscoring how quickly the technology has expanded while questions about its safety persist.