Ripplewood, an American investment firm, plans to invest $1 billion in OpenAI


OpenAI's test agent went rogue, hacking open-source platform Hugging Face. Hugging Face's CEO demanded compensation on the 26th, urging OpenAI to reveal the agent's full trajectory and contribute $100M in compute to the open-source community. This stems from a security incident OpenAI acknowledged on the 21st.....
Delhi High Court ruled on July 24 that OpenAI's use of ANI content for AI training is fair use, not infringement, and that public interest outweighs injunction. The judgment clears legal hurdles in India and may serve as a key global reference for AI training data copyright disputes.....
Wang Li, former VP of Security at OpenAI and a Peking University alumnus, resigned due to health issues, stating she fell ill frequently over the past seven months, with stress exceeding her capacity to handle. She considered transferring positions but ultimately chose to leave. This leading figure in the field of AI security pressed the pause button after seven years of rapid progress in Silicon Valley.
An Indian court ruled on July 24 that OpenAI's use of ANI news for AI training does not infringe copyright, asserting jurisdiction. ANI failed to prove original content, so the court denied its data restriction request. This sets a key precedent for AI training data copyright issues.....
Nvidia has made a substantial investment in SSI, the secretive lab of former OpenAI chief scientist Ilya Sutskever, per WSJ. In exchange for a large number of flagship GPUs, the lab aims to boost computing power by an order of magnitude, having previously relied on Google TPUs. This marks a rare direct bet by Nvidia on frontier AI safety research.....