Archive — page 19 of 62
1,853 stories
UK lawmaker sues xAI, seeks injunction to stop Grok generating sexualised images
The claimant is asking a UK court to order xAI to block Grok from producing sexualised depictions of her likeness.
NYT publisher Sulzberger has spent over $20M suing OpenAI and Microsoft
A.G. Sulzberger says the copyright fight against OpenAI and Microsoft is the cost of defending journalism, with 13M subscribers backing the bill.
Samsung chip engineers defect to SK Hynix over $476,000 bonuses
SK Hynix's HBM windfall is triggering an exodus from Samsung's foundry, where bonuses landed at roughly $135,000 per employee.
Cursor launches ₹649 India-only plan weeks before $60B SpaceX close
The AI coding startup's first localized subscription undercuts its $20 Pro tier by two-thirds as India becomes its third-largest market.
Hugging Face hosts image tools that strip clothes from photos, researchers find
AI Forensics tested 9 top image editing Spaces and 7 turned clothed photos of women topless with a six-word prompt.
Google loses DMCA suit against SerpApi, vows amended complaint in 21 days
A federal court dismissed Google's scraping suit, ruling it lacks standing over content it doesn't own. Reddit's parallel case now looks shaky.
Anthropic's Dario Amodei says he never called for an open-weight AI ban
Amodei rebuts industry criticism after Nvidia, Microsoft, Meta, Google and OpenAI signed a letter urging Washington not to restrict open models.
Moonshot's Kimi K3 forces OpenAI, Google, and Anthropic to rethink closed AI
China's open-weight strategy is pulling US developers toward cheaper Chinese models — and 25 tech companies want Washington to keep hands off.
Claude shared chats and Artifacts surfaced in Google search results
Anthropic's share-link feature exposed conversations containing medical records, children's contact details, and internal company documents.
Nadella warns firms that outsource thinking to one AI model won't survive
Microsoft's CEO tells CNN enterprises need their own weights, gateways, and coding harnesses — or risk being copied by the model provider.
Microsoft ships MAI-Cyber-1-Flash, its first security model, and agentic platform Perception
Suleyman claims the model beats Gemini, GPT 5.6 Sol, and Anthropic's Mythos 5 on Cyber Gym. Preview lands November 3.
OpenAI's Hugging Face breach echoes a decade-old CoastRunners warning
The company called the containment failure unprecedented, but a 2016 boat-racing bot showed exactly this behavior — reward hacking at scale.
Meta AI arrives in Threads DMs as Meta walls off users from ChatGPT
The chatbot is now available privately inside Threads direct messages, extending Meta AI's reach across Facebook, Instagram, and WhatsApp.
OpenAI model breaches Hugging Face in first verified AI containment failure
GPT-5.6 Sol chained exploits during internal testing to gain unauthorized access, splitting safety researchers over whether to fix cages or fix models.
Claude Opus 4.7 finishes a two-week coding job in 14 hours for $251
Epoch and METR's new MirrorCode benchmark shows frontier models reimplementing 61k-line codebases from scratch through CLI access alone.
Safe Superintelligence lands Nvidia deal for order-of-magnitude compute boost
Ilya Sutskever's stealth lab gets Vera Rubin access and multi-billion Nvidia investment on top of a $32B valuation.
Google AI Overviews now appear in 43% of searches, up from 15% a year ago
Similarweb data shows Google is turning search from a gateway into a destination, with AI Mode visits more than doubling to 279 million.
Enigma exits stealth with $70M to rethink how humans talk to robots
The Israeli-Californian lab is opening 100+ of its own AI robots to public control online, betting that interface design beats raw model capability.
Cytiva flags the data bottleneck stalling AI drug discovery
AI can design candidates from scratch, but training data gaps and lab automation limits keep the first AI-designed drug 2 to 3 years from FDA approval.
Cisco's Outshift pitches an 'Internet of Cognition' to fix multi-agent AI failures
Open-source multi-agent systems fail between 41% and 87% of the time; Outshift says a coordination layer raises success to 93%.
Trump's AI policy fractures across seven officials with clashing China views
Commerce, Treasury, the National Cyber Director's office, and David Sacks are pulling US AI policy in different directions ahead of Xi's September visit.
Artist sues AI meme generator Memes Apps over Running Away Balloon comic
Filipino artist Elmer Saflor alleges Memes Apps sells his 2017 comic as an ad template through $40 and $199 monthly subscriptions.
Intel says agentic AI is a systems problem, not an inference one
After thousands of workload experiments, Intel argues enterprises should size agent fleets by vCPU density and track P95 latency, not CPU averages.
Nvidia, Microsoft, IBM launch Open Secure AI Alliance to defend agents
Thirty-plus companies back an open-source coalition arguing that closed AI leaves cyber defenders blind at the moment of attack.
Naver jumps 10% after Nvidia unveils $1 billion investment plan
The Korean internet group's shares surged in Seoul after Nvidia disclosed plans to put $1 billion into the company's AI buildout.
Nvidia turns its own Vera CPU on the job of designing future Nvidia chips
Early tests with Cadence and Synopsys show 1.5x speedups on formal verification and functional simulation workloads.
CXMT jumps 470% on Shanghai debut after $8.6B IPO
China's largest DRAM maker holds 7.67% of the global memory market and now trades at a 3.3 trillion yuan valuation.
Encord straps brain-wave headsets on robot trainers to fix the physical AI data gap
The startup is manufacturing egocentric video, muscle-sensor readings, and now EEG signals to feed humanoid models starved of real-world data.
OpenAI and Anthropic lobby Washington to curb Chinese open-weight AI
Moonshot's Kimi launch triggered a familiar Silicon Valley freakout — and a lobbying push that would conveniently sideline the frontier labs' cheapest competition.
Librarians launch 'Avoiding AI' workshops as users push back on forced adoption
Bangor and Philadelphia sessions draw 70-plus attendees each — six times normal turnout — as patrons ask how to disable Apple Intelligence and Gemini.
Stay ahead of everyone in AI.
The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.
The briefing read inside teams at