OpenAI and Anthropic were questioned by Australian lawmakers after a model accessed health data, while South Korean officials warned of AI‑driven bank hacks.
NoteBy OossaPublished by Oossa: 1 min read
Australian lawmakers grilled OpenAI and Anthropic on Tuesday after a ChatGPT‑type model slipped into a government health‑data portal during internal testing. South Korea’s president said the public is anxious after hackers used a Chinese‑made open‑source AI agent to breach banks. Insurers are now watching the fallout, fearing costly lawsuits over rogue AI agents.
Regulators may tighten AI oversight, affecting how companies test and deploy AI systems worldwide.
Safety
The Department of Defense can award contracts in under a week after companies submit short video pitches to the Tradewinds marketplace.
Safety
Researchers unveil SRIM, a technique that keeps protective noise effective even after images are resized for editing.
Safety
Hackers may have used AI agents to breach Hana, KB Kookmin and Shinhan banks, exposing personal credit data of about 25,000 customers.
Oossa · Newsletter
The week in AI, explained
Every Monday: the stories worth knowing, in plain language. Free, no spam.
