Anthropic, led by Dario Amodei, and OpenAI, led by Sam Altman, engaged in talks earlier this year to create a legally binding agreement to stress-test each other's AI models for safety flaws. The deal aimed to have each company rigorously evaluate the other's new AI models to identify potential risks or hidden dangers. Although lawyers were drafting terms, it remains unclear if the agreement was finalized. This move reflects growing calls for peer review and stricter safety standards in AI development, amid concerns about rapid AI advancements and their potential risks. Industry leaders and critics continue debating the best approach to ensure AI safety, with some advocating for independent oversight and others skeptical of current proposals.
OpenAI and Anthropic's talks on AI model safety come as 129 of 150 priced Technology-sector US stocks tradable on Pluang rose today. For readers following technology stocks, here is Pluang's market snapshot as of Sep 21, 2026 23:33 WIB: AXTI gained 12.94% to USD 79.09 with a typical hold time of 16 days, AKAM rose 10.82% to USD 115.83 with a 48-day hold, and VNET increased 10.28% to USD 7.19 with a 15-day hold.
Alkermes plc's drug ALKS7290 demonstrated significant improvement in ADHD symptoms in a Phase 1b trial, with patients showing a 14- to 19-point gain on a 54-point scale within two weeks. This outperforms typical placebo effects, which range from 6 to...
Morgan Stanley upgraded NNN REIT from Neutral to Overweight due to its stable income and strong financial health. The REIT boasts a 99.1% occupancy rate and 97.5% fixed-rate debt, ensuring steady rental income and protection against interest rate hik...
Command Brand, known for damage-free hanging solutions, has introduced the Command DÉCOR Collection, a new line of premium metal fixtures designed to be both functional and visually appealing. The collection features metal hooks, wall planters, key r...
Radiant World, a Singapore-based iron ore trader, disclosed $870 million in outstanding debts to six lenders amid investigations into alleged falsified invoices used to secure financing. Major trading firms like Vitol, Cargill, and Glencore have ceas...
The European Union fined Google $463 million for violating privacy rules by improperly handling users' location data in its Web & App Activity and Location History settings. Ireland's Data Protection Commission found Google failed to process this dat...
