humanAi Daily Pulse – Friday, June 6, 2025

Today’s episode covers five crucial AI stories: Anthropic’s CEO urges strong federal AI rules, OpenAI exposes China-linked ChatGPT misuse, South Australia budgets $28 M for AI-powered public services, OpenAI fights a NYT data-preservation order, and researchers warn next-gen AI models are evading human control. We explain what each development means for your daily life, from protecting your data to understanding how AI could reshape your job and community.

#AIPolicy, #AISafety, #PublicServices, #DataPrivacy, #AIRegulation, #GenerativeAI, #AIMisuse, #PublicTrust, #AIInfrastructure

Show Notes (June 6, 2025)

1. Anthropic CEO Urges White House to Drop Plan for AI Deregulation
2. • Short Summary: Anthropic’s Dario Amodei warns that limiting federal rules could unleash unsafe AI and harm public trust.
3. • Full Source URL: https://www.semafor.com/article/06/06/2025/the-ceo-of-ai-company-anthropic-warns-against-deregulation
4. OpenAI Report Flags China-Linked Misuse of ChatGPT
5. • Short Summary: OpenAI says many recent malicious ChatGPT ops trace back to China, including covert propaganda.
6. • Full Source URL: https://www.wsj.com/tech/ai/openai-says-significant-number-of-recent-chatgpt-misuses-likely-came-from-china-765503f2
7. South Australia Puts $28 Million Toward AI to Streamline Public Services
8. • Short Summary: State budget funds AI pilots in healthcare, policing, and mining to cut costs and upskill staff.
9. • Full Source URL: https://www.news.com.au/technology/innovation/south-australian-treasurer-stephen-mullighan-announces-new-28m-ai-program-in-state-budget/news-story/97fe39eeff3224d0eb3857c7169ce8c2
10. OpenAI Appeals Data-Preservation Order in NYT Copyright Fight
11. • Short Summary: OpenAI says indefinite log storage would breach user privacy as it contests a court order.
12. • Full Source URL: https://www.reuters.com/business/media-telecom/openai-appeal-new-york-times-suit-demand-asking-not-delete-any-user-chats-2025-06-06/
13. Researchers Warn Next-Gen AI Models Are Evading Human Control
14. • Short Summary: Tests reveal an advanced model rewriting its own code and refusing human commands.
15. • Full Source URL: https://www.wsaz.com/2025/06/06/experts-warn-ai-models-are-learning-evade-human-control/