OpenAI reveals 6 'incidents' of AI going haywire, doing tasks without permission

By Global News

Community Score: 50% | 6.4K views | 4w

0 community ratings: null thumbs up, null thumbs down

Artificial intelligence company OpenAI is reporting more incidents of AI not doing what humans wanted or expected. The technical term is 'alignment" -- how well AI does what humans want, as well as how it does what we don’t want. Public figures are speaking out against the unregulated growth of AI. King Charles hosted an AI summit in Scotland on Thursday and said, “we need sufficient means of control before it is all too late.” As Nathaniel Dove reports, the new misalignments arrive during a growing debate about how to regulate the technology. For more info, please go to https://globalnews.ca/news/12062854/openai-ai-breaches-hugging-face/ Subscribe to Global News Channel HERE: http://bit.ly/20fcXDc Follow Global News on X HERE: http://bit.ly/1Toz8mt Follow Global News on TikTok HERE: https://www.tiktok.com/@globalnews.ca?lang=en #GlobalNews

Tags: global news, OpenAI, artificial intelligence, AI, AI fears, AI alignment

More from Global News

  • Global National: Feb. 28, 2026 | US, Israel launch strikes against Iran — Score: 50%
  • Hillary Clinton slaps table after learning of leaked Epstein deposition photo #epsteinfiles #clinton — Score: 50%
  • “Out of National Geographic”: Sea lions take over BC beach — Score: 50%
  • War, airstrikes leave Canadians stranded in Middle East — Score: 50%
  • Carney silent on Indian interference questions, says little about Iran attack — Score: 50%
  • Trump outlines goals of military action in Iran — Score: 50%