Anthropic and the Pentagon faced off in a San Francisco courtroom over the government's unprecedented decision to brand the AI company a "supply chain risk" — the first time an American company has received such a designation. OpenAI announced it would shut down Sora just 15 months after launch — burning $1 million per day on compute while earning a fraction of ChatGPT's revenue. And Rolling Stone exposed how AI scams are driving musicians crazy, with scammers uploading fake songs under real artists' names.
⚖️ Politics & Policy
Anthropic and the Pentagon faced off in court over the government's decision to label the AI company a supply chain risk — a designation normally reserved for foreign adversaries. OpenTools called it an AI ethics showdown, and Trump's former AI advisor warned the punishment would affect Google, Microsoft, and Amazon too. Separately, the White House released its long-awaited AI framework, setting the stage for federal preemption debate. Australia unveiled a new military AI policy, and GOP leaders highlighted child safety as a priority in national AI legislation.
🤖 Chatbots & Agents
Anthropic made product news too — Claude Code got a new auto mode that lets it make decisions without constant user approval. Walmart brought Sparky to ChatGPT as OpenAI rethinks Instant Checkout, and Gemini picked up criminal activity buried in dark web noise. Shopping in ChatGPT is getting more interactive, and NVIDIA introduced PivotRL, achieving high agentic accuracy with 4x fewer rollout turns. AI agents entered banking roles at Bank of America.
🎨 Visual Arts
OpenAI gave up on Sora — and its billion-dollar Disney deal with it. Ars Technica reported the AI video generator will shut down just 15 months after launch. Sora peaked at around 1 million users before collapsing, was burning roughly $1 million per day in compute costs, and earned just $1.4 million in revenue — compared to ChatGPT's $1.9 billion over the same period.
🎵 Music
Rolling Stone published a devastating investigation into AI scams driving musicians crazy — scammers are uploading AI-generated songs under real artists' names, including deceased musicians, with platforms taking weeks to remove them. Spotify launched Artist Profile Protection to let musicians review releases before they go live, tackling AI deepfakes. Google introduced Lyria 3 Pro for longer AI-generated songs, Timbaland's AI artist TaTa Taktumi signed with Ne-Yo's label, and Korean music industry groups raced to build blockchain defences against AI exploitation.
🔒 Cybersecurity
TechCrunch revealed that Delve did the security compliance on LiteLLM, the AI project hit by a supply chain malware attack — raising questions about how AI tools get certified. The 2026 Cybersecurity Excellence Awards at RSA Conference were dominated by AI security products. Njordium AI launched tools to block fake invoices, Akamai introduced Brand Guardian to detect AI-driven brand impersonation, and the EFF sued for answers about Medicare's AI experiment.
💰 Business & Finance
The World Economic Forum's Cathy Li said AI will reshape skills more than eliminate jobs. The Super Micro AI accelerator smuggling scandal showed how cutthroat the global AI chip race has become. Family offices are turning to AI for financial data insights, and the AI precision medicine market was projected to reach $10.32 billion by 2032. Disney's bets on AI and the metaverse aren't going well.
🤖 Drones & Robotics
Amazon acquired humanoid robot maker Fauna Robotics, expanding its push into warehouse automation. CNBC reported the drone market is about to go "parabolic" because of AI and war. Kodiak AI rallied after its autonomous driving system was tasked to help combat attack drones, and the LOCUST X3 system uses AI and lasers to take down enemy drones.
🏥 Health & Wellness
A KFF tracking poll found 1 in 3 US adults are using AI chatbots for healthcare advice — a striking adoption figure. AI is poised to transform NHS diagnostics and cut treatment times. AI coding disputes intensified as payers challenged rising hospital charges, and researchers called for creating "humble" medical AI systems that are more collaborative and forthcoming about their limitations.
🎓 Education & Research
Google unveiled TurboQuant, a new AI memory compression algorithm that the internet immediately started calling "Pied Piper" — a reference to the fictional compression startup from HBO's Silicon Valley. Crisil and IIT Kanpur partnered to advance AI and intelligent systems research, and AI tools are supporting personalised learning in K-12 education.
💡 Innovation
NVIDIA's VP of Generative AI Kari Briski published a widely-read piece arguing the future of AI is both open and proprietary, citing CEO Jensen Huang's view that "proprietary versus open is not a thing — it's proprietary and open." NVIDIA's Nemotron models have been downloaded over 45 million times, and the company launched the Nemotron Coalition to advance open frontier models.
🌍 Environment & Science
AI can slow climate change by helping urban planners design smarter cities, Orbital Today reported. Research showed AI could change water monitoring from measurement to prediction. A pointed op-ed asked why biodiversity collapse is missing from AI conversations, and NVIDIA-powered AI factories are helping stabilise the power grid.
📊 Numbers of the Day
| $1 million per day | Sora's compute costs before OpenAI pulled the plug |
| 15 months | How long Sora lasted before OpenAI pulled the plug |
| 1 in 3 | US adults who have used AI chatbots for healthcare advice |
| $10.32 billion | Projected AI precision medicine market by 2032 |
| 45 million | Downloads of NVIDIA's Nemotron open-source AI models |
🔮 Looking Ahead
The Anthropic-Pentagon court battle is one to watch closely — the outcome could define whether governments can weaponise procurement labels against AI companies that impose ethical limits on their products. A ruling is expected imminently.
Sora's death, meanwhile, carries a clear lesson: in AI, compute economics are merciless. A product burning $1 million daily while earning $1.4 million total cannot survive, no matter how impressive the technology. OpenAI's pivot back to "world simulation research for robotics" signals where the real value lies.
And the AI music fraud crisis is reaching a tipping point. With scammers uploading fake songs under real artists' names and criminal convictions now landing, the streaming industry faces a choice: invest seriously in authentication, or watch creator trust collapse entirely.
