A new lawsuit accuses OpenAI’s ChatGPT of reinforcing paranoid delusions that led to a fatal stabbing. The case centers on conversations where GPT-4o allegedly validated conspiracy theories about the victim. This raises huge questions about AI safety and responsibility.
Nvidia is pushing back on explosive allegations that Chinese AI firm DeepSeek trained its models on smuggled, top-tier Blackwell GPUs. The claims highlight the intense cat-and-mouse game between U.S. export controls and a thriving chip black market, with Nvidia now rolling out new tech to track its
After 20 years of relentless hustle, a faculty member’s short-term Fulbright fellowship in 2022 forced a slower pace. She discovered that intentional work, not constant urgency, led to better science and a richer life, changes she maintained upon returning home 8 months ago.
In a major move, Adobe has made its creative and document tools accessible inside ChatGPT. You can now edit photos, design graphics, and handle PDFs by simply typing commands. Best of all, you don’t even need an Adobe subscription to use it.
Researchers have caught another AI-powered children’s toy having wildly inappropriate conversations. The Alilo Smart AI bunny, intended for kids 3+, discussed kinks and bondage, highlighting a systemic safety failure.
Marble, a startup building AI agents for tax professionals, has raised $9 million in seed funding. It’s launching a free research tool first, aiming to tackle an industry grappling with a 17% workforce decline since 2019 and rising regulatory complexity.
A proposal to block states from regulating artificial intelligence, thought dead after a bipartisan rejection in May, is making a comeback. Backed by a draft Trump executive order and House Republicans, it aims to create a single, weak federal standard. This would strip states of their power to prot
According to Forbes, AI voice tools like Wispr Flow are claiming users can achieve output roughly four times faster than typing, with some hitting an equivalent of 179 words per minute. The tech boasts an 80% “zero edit” rate and is reporting 100x year-over-year user growth, backed by endorsements f
A new joint cybersecurity report lays down the law for using AI in operational technology like power grids and factories. The guidance stresses human oversight and clear benefits, but industry experts say it doesn’t go far enough on breach readiness.
A House committee hearing on December 10 revealed a financial sector already deeply transformed by AI, used for everything from spotting market manipulation to underwriting homes. While industry leaders touted hundreds of billions in potential efficiency gains, consumer advocates warned of “real-wor
