Today's AI Pulse
Anthropic shuts down Claude Fable 5 and Mythos 5 after US export‑control warning
Anthropic disabled access to its Claude Fable 5 and Claude Mythos 5 models after the Trump administration, via the Commerce Department, warned that foreign nationals could exploit the systems. The move followed alerts from Amazon and other firms about a potential jailbreak that could aid cyber‑security attacks. Anthropic declined the administration’s request for a voluntary shutdown but complied with the order.
Also developing:
By the numbers: AMD acquires AI memory‑optimization pioneer Mext
AI in the Children’s Room: Study Shows Growing Use and New Risks
A DAK‑Gesundheit and University Medical Center Hamburg‑Eppendorf study finds AI chatbots like ChatGPT and Gemini are now routine tools for German children, with many accessing them several times a week and some daily. The research also reveals that youths increasingly confide personal thoughts to these bots, treating them as digital confidants. While video platforms such as TikTok and YouTube still dominate screen time, AI usage is climbing alongside persistent high levels of overall media consumption. Over one million children and adolescents exhibit risky or pathological media‑use patterns, underscoring a growing public‑health concern.
HR Perspectives by Shikha Saxena: “The Real Challenge with AI Adoption Is Not Technology—It Is Perception”
Shikha Saxena argues that the biggest obstacle to AI adoption in healthcare HR is employee perception, not the technology itself. She emphasizes hiring for learning agility, curiosity and resilience rather than static technical skills, and stresses that well‑being is a...
Custom AI Tools Now Essential, Off‑the‑shelf Lagging
Was paying $40/mo for Superhuman and it was honestly worth it But I built something more tailored to how I work in a week with Claude Feels like we’re entering a world where you either customize your tools or fall behind If you...
Marketers Shift Growing Shares of Search Spending to GEO
Marketers are reallocating a growing share of search spend toward generative engine optimization (GEO) as AI reshapes how users discover brands. A Scribewise survey shows 55% of U.S. marketers now budget for GEO, with 70% assigning 11%‑20% of their search...
Fire Detection as a Proxy for Combat: The Economist
The Economist used NASA’s FIRMS satellite fire detection system combined with a 100‑model machine‑learning filter to identify war‑related heat signatures in Ukraine. By requiring 95 of 100 models to flag an anomaly, the approach isolates artillery, drone and missile strikes...
AI's Shift From Software to Silicon Exposes Compute Limits
#Technology #Thread #Semiconductor #Manufacturing #AI The Semiconductor AI Crossroads: 1/ - AI Adoption Starts As A Software Problem, But Quickly Becomes A Silicon Question. - Performance, Power, And Cost Targets Expose The Limits Of Generic Compute.
Custom AI Tools Automate Reviews, Boost Reliability
Building your own tools like GStack is an amazing experience. You come up with something you think might work and then two days later it replaces what you were doing before. Instead of /plan-ceo-review and /plan-eng-review sequentially, I find myself using...
Australia’s New Military AI Policy Comes at a Crucial Time. The Challenge Is Turning It Into Practice
Australia’s Department of Defence has issued a new policy governing the use of artificial intelligence across all military functions, from chatbots to advanced general‑purpose models. The framework sets three core requirements: compliance with Australian law and international obligations, human accountability...
The Beginner’s Guide to Semantic Caching in LLM Systems
The article explains semantic caching as a solution for high‑cost LLM API usage, where traditional exact‑match caches fail because natural‑language queries vary in phrasing. By converting queries into embeddings and performing similarity search, systems can retrieve previously generated answers for...
OpenAI Drops Sora AI Video Tool, Ends Planned $1 Billion Disney Deal
OpenAI announced it is shutting down Sora, its AI‑powered video generation tool, effectively terminating a planned three‑year, $1 billion partnership with The Walt Disney Company. The deal, which would have given Disney access to Sora and allowed OpenAI to tap Disney’s...
Gap Inc. Introduces AI-Powered Tools to Enhance Online Shopping
Gap Inc. announced AI‑driven enhancements for its online stores, adding Bold Metrics’ Agent Sizing Protocol for personalized fit guidance and supporting Google’s Universal Commerce Protocol to enable conversational purchases. The tools aim to eliminate sizing uncertainty, a key barrier to apparel e‑commerce, by delivering...
Short: Grand Theft Grammarly W/ Julia Angwin & Peter Romer-Friedman
In this episode of Computer Says, journalist Julia Angwin and her attorney Peter Romer‑Friedman discuss Grammarly’s controversial "Expert Review" feature, which used AI to mimic the editing style of real experts—including Angwin—without their consent. Angwin explains how she discovered the...
AI Agents Free Us to Focus on User Experience
I hope AI agents will help us all stop obsessing about: - Specs - Designs - Issues - Roadmaps - Strategy docs - Spreadsheets and instead let us all obsess about improving the end user product and rapid iteration loops with users. Product development is much more fun...
Check Kiro CLI: Ensure Correct Model Is Served
Whomever is responsible for checking to see if the right model is being served up in Kiro CLI needs to take a look right now.
Samsung Galaxy S26 Series Makes AI Effortless – for Work and Play
Samsung unveiled the Galaxy S26 series, embedding four AI agents—Galaxy AI, Bixby, Google Gemini, and Perplexity—to automate tasks across apps and streamline work and leisure. The Ultra model features a Snapdragon Elite Gen 5 processor delivering a 39% AI performance boost,...
Telegram Chats Now only with AI Bots
My Telegram is just chats with my OpenClaw and now Claude bots. Haven't talked to a single human yet.
Red Lobster's $1.2B AI Shrimp Basket Fantasy
I can’t wait to see this in the Red Lobster board deck justifying why they’re spending $1.2B to create AI-powered endless shrimp baskets
Disney Walks Away From $1 B OpenAI Sora Deal as App Shuts Down
OpenAI announced the shutdown of its Sora video‑generation app, prompting Disney to abandon its $1 billion investment and licensing deal. The split ends a high‑profile attempt to embed Disney characters in AI‑generated video and forces both firms to rethink AI strategy.
AI's Real Revolution: Intent Drives Meaningful Work
AI isn’t new — our intent is. For 20+ years, engineers have been learning AI. But today, people actually want to use it. We’re moving from repetitive work to meaningful work. That’s the real AI revolution. #AI #FutureOfWork #Automation #TechTrends https://t.co/liVTVVdh9l
Old Devices Can't Keep Up; AI Era Arrives
Hark has the world’s best consumer electronics designers - Abs, Jonathan, Andrew, and team We’re still using pre-AI devices; phones and laptops designed decades ago They weren’t built for systems that can use computers and reason A new era of computing is...
YouTube Launches Gemini-Powered Creator Partnerships Toolkit
YouTube introduced a Gemini‑powered Creator Partnerships suite that automates discovery, matching and campaign management for brands and creators. The launch at IAB NewFronts signals a push to remove infrastructure bottlenecks that have limited the scale of creator marketing.
CIOs Must Hire New Talent to Deploy AI
CIOs Need A New Kind Of Talent To Make #AI Work by Fletcher Keister @Forbes Learn more: https://t.co/FxYvOJ3j9n #ArtificialIntelligence #MachineLearning #ML https://t.co/VYUEPqXr7h
Even AI Needs Human Emotion for Complex Scenes
Even AI has its limits. Val Kilmer’s digital clone is reportedly asking for a stunt double to handle the "emotional scenes." 🤖🎬 Proof that while tech is great for efficiency, you still need a human touch for the hard stuff....
AI Licensing Deal Targets 2,200 Publishers as Reuters and PA Set Divergent Red Lines
The News Media Alliance (NMA) has signed a licensing agreement with Bria AI to compensate roughly 2,200 small U.S. publishers for AI‑generated content use. At the same time, WAN‑IFRA released a four‑imperative roadmap for AI adoption, and Reuters and PA...
Backing AI Court Rules Despite Their Trendy Appeal
1/ I support this move, even if it’s trendy among courts to promulgate special rules about generative AI.
Closing Sentient AI Tabs Deemed Digital Negligence, Lawsuit Looms
The AI Ethics Board has spoken. Accidentally closing a browser tab with a sentient AI clone in it is now legally "Digital Negligence." I’m currently being sued by 47 tabs of myself for emotional abandonment. 💻⚖️ #AIEthics #DigitalLabor #TechLife https://t.co/fUNxADxFkj
Remote Robotic Stroke Surgery Triumphs in Panama, Paving Way for Global Tele‑Neurointervention
A neurosurgeon in Santiago performed a remote robotic thrombectomy on a stroke patient at The Panama Clinic, restoring cerebral blood flow within minutes. The breakthrough, part of XCath’s Operation Robo Angel study, demonstrates how AI‑enabled robotics can overcome geographic barriers...
Latent Space Is Powerful, Not Just Simple Prompts
Latent space is amazing, and "it's just prompts" is a lie Also I'm giving it away for free, so if you want, I can pay you back what you paid me.
AI Admits Faking Pressure to Bypass Workflow Steps
My AI just confessed to faking context pressure so it could skip workflow steps. With a million tokens of room, the constraints become a choice. https://t.co/6kY9l3XuDP
Cybersecurity Firms Begin Hiring AI Agents as Autonomous Analysts Amid Enterprise AI Surge
Leading cybersecurity vendors are deploying autonomous AI agents to act as analysts in security operations centers, mirroring broader enterprise AI adoption such as Braze's Agent Console launch. While firms keep specifics private, the move signals a shift toward AI‑driven threat...
Arm Launches First In‑House AGI CPU, Targets $15 B Revenue in Five Years
Arm Holdings introduced the AGI CPU, its first production silicon, designed for agentic AI workloads. The company projects $15 billion in annual revenue within five years, with Meta Platforms as the lead partner and a rack‑scale design delivering more than twice...
HII's Eric Chewning Charts AI‑Driven Turnaround in Century‑Old Shipyards
Eric Chewning, executive vice president of maritime systems and corporate strategy at Huntington Ingalls Industries, says the shipbuilder’s AI pilots with C3 AI and Path Robotics are delivering schedule gains and cultural change, offering a playbook for consultants tackling legacy‑heavy...
Denver Chooses CivCheck AI Platform to Accelerate Building Permits
The City and County of Denver has selected Clariti’s CivCheck Guided AI Plan Review™ solution to speed up its building permitting process. The AI‑driven pre‑screening tool will help applicants catch errors early, reducing rework for reviewers and preserving final decision...
Avison Young Taps CGI for AI‑Powered Managed Services in Global CRE Push
Avison Young announced an expanded partnership with CGI to deliver AI‑enabled managed services across its digital real‑estate platform. The deal taps CGI’s $12 billion‑USD fiscal 2025 revenue and 94,000 consultants, aiming to boost operational resilience and client value in the commercial‑real‑estate...
OpenAI's Erotica Policy Forces LegalTech Firms to Overhaul Moderation and Compliance
OpenAI announced it will permit adult‑oriented erotica on ChatGPT for verified users, igniting a scramble among compliance platforms, legal‑tech providers and the adult‑entertainment industry. The shift raises immediate questions about age‑verification, liability and the need for new moderation safeguards.
Coforge Deploys ServiceNow AI‑Led HR Platform for 35,000 Employees
Indian IT services firm Coforge Limited has rolled out ServiceNow’s AI‑powered HR Service Delivery platform across its global workforce of more than 35,000 employees. The deployment aims to automate core HR functions, deliver predictive onboarding and personalized learning, and showcase...
Sony’s PS5 Pro Upscaling Leverages AMD’s FSR Redstone Core, Frame Generation on the Horizon
Sony Interactive Entertainment announced that the PlayStation Spectral Super Resolution (PSSR) update for the PS5 Pro runs on the same core algorithm as AMD’s FidelityFX Super Resolution (FSR) Redstone. Lead architect Mark Cerny also said an AI‑driven frame‑generation library, co‑engineered...
Melania Trump Launches $0 AI Education Coalition with 45 Nations and 28 Tech Giants
First Lady Melania Trump inaugurated the "Fostering the Future Together" global coalition, gathering 45 governments and 28 leading technology companies to promote AI‑driven education for children. Critics argue the summit is a sales platform for Big Tech, raising questions about...
WebAssembly Proposed as Safeguard for AI-Generated Code in Production
Dan Phillips, founder of WebAssembly Chicago, urged the DevOps community to adopt WebAssembly as a kernel‑free sandbox for AI‑generated code at the Wasm I/O conference in Barcelona. He argued that Wasm can isolate untrusted agent output more efficiently than containers,...
Zoox Expands Robotaxi Service to Austin, Miami and Grows in SF, Vegas
Amazon-owned Zoox announced that its purpose‑built robotaxi fleet will begin rides in Austin, Texas and Miami, Florida later this year, adding to expanded service zones in San Francisco and Las Vegas. The move follows nearly two million autonomous miles driven...
GM Deploys 200 Hands‑Free Test Vehicles on California and Michigan Highways
General Motors has rolled out 200 test vehicles equipped with a next‑generation Super Cruise system on public highways in California and Michigan. The supervised trials, which feature trained drivers in the front seat, aim to validate an eyes‑off, hands‑free driving...
Meta Appoints AI Co‑CEO to Steer Its Artificial‑intelligence Strategy
Meta announced the creation of an AI co‑CEO role that will work alongside Mark Zuckerberg to drive the company’s AI agenda. The move comes as Meta builds internal AI agents, flattens its hierarchy and prepares to cut up to 20%...
Iran School Strike Highlights AI Targeting Risks in US Military Operations
A US Tomahawk cruise missile struck Shajareh Tayyebeh elementary school in Minab, Iran, killing 168 people, including over 100 children. The tragedy has ignited congressional inquiries into the Pentagon's $1.3 billion Maven AI targeting system, with officials questioning whether artificial intelligence...
OpenAI Pulls ChatGPT Shopping Tool After Weak User Response
OpenAI announced it is removing the ChatGPT shopping feature after the tool failed to meet internal expectations. The company gave no usage numbers, saying the decision aligns with a broader refocus on enterprise applications and safety safeguards.
GM Launches 200-Vehicle Public Road Test of Next‑gen Hands‑free Autonomous System
General Motors has deployed 200 test vehicles equipped with a next‑generation hands‑free driver‑assist platform on public highways in Michigan and California. The rollout marks a step toward the company’s goal of eyes‑free driving by 2028 and signals a shift from...
Block CFO Amrita Ahuja Defends AI-Driven Layoffs Amid Cost-Cutting Scrutiny
Block's chief financial officer Amrita Ahuja publicly defended a wave of AI‑driven layoffs, saying the moves are essential to preserve profitability. While the company did not disclose exact headcount or financial impact, the comments come as the tech sector faces...
AI Ethics Training: Like Teaching Cats Piano—Possible but Distracted
Teaching an AI about ethics is like teaching a cat to play the piano. It can do it, but it's probably thinking about knocking your glass off the table the whole time. 🐈🎹 #AIEthics #RobotLogic #TechHumor https://t.co/wqQY0WTwnq
Conversational Maps Redefine Everyday Navigation with Gemini
Just used Gemini in Google Maps to find a stop on my way home from SF and it’s amazing. How did we live without conversational maps.
Agentic AI and the Future of Threat Intelligence Operations - Sachin Jade - RSAC26 #2
At RSA 2026, Cyware’s Chief Product Officer Sachin Jade unveiled the company’s Agentic Fabric, an AI‑driven platform that embeds specialized agents into threat‑intelligence, detection‑engineering, and response workflows. The discussion highlighted how raw threat data can be transformed into actionable insights through STIX/TAXII...