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
UF’s HiPerGator Drives FNN’s AI Wildfire Detection in Verizon Prize Recognition
A Gainesville startup founded by University of Florida alumni, FNN, secured a share of Verizon’s $1 million Disaster Resilience Prize. Leveraging UF’s HiPerGator supercomputer, FNN’s AI platform shrinks wildfire detection time from 24 hours to 40 seconds and adds real‑time smoke monitoring. The prize will fund scaling of its lightning‑sensor network across fire‑prone regions, accelerating deployment for utilities and fire agencies. The win highlights the synergy between academic HPC resources and commercial disaster‑response technology.
More Foreign Visitors Meet Xi; Service Sector Boost; Countering Improper Foreign Conduct; Youth Unemployment; AI Chip Shortage; More Pressure on...
Chinese President Xi met Mozambique President Daniel Chapo and Lao envoy Saleumxay Kommasith, stressing China’s readiness to help Africa navigate the spillover of Middle East conflicts and upgrading bilateral ties with Mozambique to a community‑level partnership. Meanwhile, former U.S. President Donald Trump...
A Student's Path to Publishing
Alessa Carbo, a junior computer‑science student at Johns Hopkins, co‑authored a paper on AI‑powered translation of sign‑language videos that was accepted at the 2025 EMNLP conference. The work, developed during the Frederick Jelinek Summer Workshop, uses a custom vision‑language model...
Prepare for Agentic AI: When Agents Talk to Agents
The question isn't if you'll use Agentic AI—it’s how you’ll handle it when your customers’ AI agents start talking to your AI agents. 🤖💬 91% of leaders are already seeing AI-driven crime. The "manual" era is over. Join @twifintech x @Alloy on...
KIME Review 2026: The AI Search Visibility Platform Built for Brands That Want to Win in LLM Search
KIME is a purpose‑built AI search visibility platform that monitors how Shopify brands appear in large‑language‑model responses such as ChatGPT and Perplexity. It provides prompt‑level share‑of‑voice, citation analytics, sentiment scores and an Action Centre that delivers weekly optimization recommendations. Pricing...
Chegg First to Crumble; Which Large‑Cap Faces AI?
Chegg. The official first company to get decimated by AI. What large cap stock do you think has the ability to have this happen due to AI?
Our Curated List of AI's Must‑Watch Developments
What is really worth your attention in the busy, buzzy world of AI? Our reporters and editors have spent years thinking about this question, charting AI’s progress and mapping out what’s next. Now, for the first time, we’ve distilled our...
How to Move Beyond the AI Pilot
Companies are flooding the market with AI pilots, yet most never leave “pilot purgatory.” Research shows 80% of firms run pilots but only 5% achieve enterprise‑wide value, despite impressive metrics such as a $2.3 million inventory‑optimization saving. The article proposes a...
UnitedHealth Group Investing Billions in AI
UnitedHealth Group announced a $1.5 billion investment in artificial intelligence to modernize its internal operations and launch commercial AI products. The spend targets member and patient engagement tools, workflow automation, and productivity gains across its health‑care and insurance divisions. In Q1...
KPMG Plans to Hand Routine Testing Off to AI
KPMG announced a multi‑year rollout of artificial‑intelligence agents to perform routine audit testing, beginning with a pilot this summer and aiming for full deployment on selected tests by 2029. The firm calls the AI “orchestration agents,” which can manage dozens...
ChatGPT Is so 2025 — Here Are the Real AI Gold Mines for Investors in 2026
AI capital is surging, with $73.1 bn raised in Q4 2025—58% of global VC funding—and a $110 bn round for OpenAI in early 2026. The United States remains the dominant spender, pledging $328 bn in government AI budgets, while Chinese firms launch cheaper large‑language...
The Oreo Cookie Method: How to Get the Best of Human Expertise and AI in Contract Review
The Oreo Cookie Method is a five‑step, tool‑agnostic workflow that places two attorney‑led reviews around AI‑driven redlining. By having lawyers set context, direct AI playbooks, run general reviews, and perform targeted clean‑ups, the process captures AI speed while preserving human...
Field Notes From Transform 2026: What Leaders Are Figuring Out About AI
At Transform 2026, HR leaders confronted the gap between AI hype and day‑to‑day execution, asking what concrete changes their teams should make this week. The conference highlighted that “AI fluency” is often a vague mandate without clear expectations, leaving managers...
Anthropic Likely a Short‑term Edge, Not Forever
Shame that these companies aren’t publicly traded. It would be a good voting machine for which AI companies are going to win. There is a hypothesis being tested: “Is Anthropic a temporary winner with a slightly better coding model?” Or “is...
Redefining Jobs for the Intelligent Age
A Kyndryl 2025 Readiness Report finds 87% of business leaders expect AI to reshape jobs within a year, yet only 29% feel their workforce is prepared. To bridge this gap, Kyndryl has partnered with Wayne State University to launch the...
This AI Warning Is A Myth; The Danger Is Not...
A new study by Carnegie Mellon, Oxford, MIT and UCLA finds that brief AI assistance erodes people’s willingness to tackle problems, even after the tool is removed. Participants using AI skipped nearly twice as many questions as a control group,...
How Zero Networks Is Closing the Network Enforcement Gap for AI Agents
Zero Networks, founded in 2019, offers an agentless, automated microsegmentation platform that eliminates manual policy creation. The solution discovers assets via directories and third‑party tools, then enforces label‑based policies using native firewalls and switch ACLs. Its new AI Segmentation feature...
Manager Of 9,000 Apartments Testing New AI Tool To Find Tenants
Bonaventure, which oversees 9,001 apartments worth $2.8 billion in assets, is piloting an AI-driven marketing platform from startup CorridorIQ. The tool aggregates 15‑20 data points—such as building permits, search trends, and mortgage applications—to forecast migration patterns and target prospective renters more...
Canada’s AI Regulation Will Be “Airtight” On Bias, Racism, and Hate, Solomon Says
Canada’s AI Minister Evan Solomon announced that the nation’s upcoming AI regulation will be “airtight” on bias, racism and hate, while remaining “light” where innovation is needed. The refreshed AI strategy, originally slated for release by the end of 2025,...
Google’s New Deep Research and Deep Research Max Agents Can Search the Web and Your Private Data
Google unveiled two new autonomous research agents—Deep Research and Deep Research Max—built on the Gemini 3.1 Pro model. The agents can combine open‑web searches with private enterprise data via a single API call, generate native charts and infographics, and connect...
AI Compute Shortage Challenges ‘Bubble’ Narrative
AI compute capacity is tightening as demand surges, undermining claims of an AI bubble. Senior director Vlad Galabov of Omdia highlighted the shortage at Data Center World. OpenAI had to scale back its video‑generation tool Sora to reallocate compute to...
Arm Exec: New AGI CPU Has Big On-Prem Potential—But Limited Channel Play For Now
Arm unveiled its first silicon product, the AGI CPU, aimed at on‑prem AI data centers handling agentic workloads. The 136‑core, 300‑watt processor promises more than double the rack performance of traditional x86 CPUs and could save up to $10 billion per...
Managing How GenAI Impacts Customer Experiences
Consumers now expect instant answers from GenAI tools like ChatGPT, prompting many firms to deploy public chatbots. However, employees in customer‑facing roles are increasingly using the same tools, risking reliance on unvetted, hallucinated information that can erode brand trust and...
ORNL’s Frontier Supercomputer Trains AI to Model Cosmic Storms
Researchers leveraged ORNL’s Frontier supercomputer—capable of 2 exaflops—to train a two‑stage AI system that captures magnetohydrodynamic turbulence in plasma with unprecedented fidelity. The hybrid model pairs a physics‑informed neural operator with a score‑based diffusion generator, halving prediction errors and delivering results...
AI Lab Aims to Automate Its Own Creation
A new AI lab built from the ground up around automated research. Led by the legend Jerry Tworek, former VP of Research at OpenAI, where he led the development of reasoning models. Absolutely incredible founding team. Heavy hitters out of OAI,...
BlackRock Launches RockAI, No‑code AI for 5,000 Developers
JUST IN: BlackRock rolls out "RockAI" to 5,000 in-house developers, a no-code platform for building custom AI agents
Virgin Atlantic Launches First-of-Its-Kind Airline App in ChatGPT to Revolutionise Travel Planning
Virgin Atlantic has become the first airline to launch an app inside OpenAI’s ChatGPT, enabling customers to search and book flights using everyday language. The AI‑driven interface instantly surfaces tailored flight options and then redirects users to the airline’s website...
AI Transforms Governance of Enterprise Unstructured Data
"Unstructured data now makes up the vast majority of enterprise information, and AI is redefining how organizations bring control, accessibility, and security to it." #DataGovernance #AI https://t.co/PYomJYHDkY
More AI Usage Beats Perfect AI Strategy
It doesn’t matter what you do with AI, just do more of it. 😅 https://t.co/eZc4HEI1d8
Montana's Frontier Learning Lab Incorporates Emerging Tech Into Education
Montana’s Digital Academy launched the Frontier Learning Lab to integrate emerging technologies such as AI and virtual reality into K‑12 classrooms. The lab’s inaugural cohort trained 20 teachers in AI, and it operates an AI help desk recognized by Google,...
Adobe AI Consumption Jumps 45% Quarter‑over‑Quarter
$ADBE shares its AI growth metrics. The 45% increase in generative AI consumption Q/Q signals enterprises are using more Adobe AI, not less, despite the Claude kills everything narrative. 👀 https://t.co/BDWlQPDn8z
AI-Driven Pushpaganda Tricks Users with Malicious Notifications
Pushpaganda uses AI-generated content and deceptive tactics to trick users into enabling notifications that deliver scams and evade traditional security defenses. https://t.co/x8sjalWKPR
How to Spot AI Audiobooks on Libby
Libby, the library e‑reading app, now labels AI‑generated audiobook narrations as “synthesized voice” or “digital voice,” allowing users to identify synthetic narrators. The article explains how to locate these tags in the NARRATOR section and how to search for them,...
New Tech Fears Echo Past Revolutions, Ignore Doomers
Everything people are saying about AI today were said about electricity, tractors, the assembly line, the ATM machine etc. Ignore the Doomers
SAP's AI Teams Blend Support with Value‑Engineering Sales
SAP's AI adoption teams: genuine support or a sales tactic in disguise? For complex products like SAP, this 'value engineering' approach helps customers navigate features and unlock more value, especially for underutilized legacy systems. #SAP #AI #CustomerSuccess https://t.co/nfoo4U8z3w
Smart Toilets Turn Waste Into Real‑Time Health Data, Launching at $200‑$400
Kohler Health, Withings and Throne Science have rolled out AI‑driven smart toilet devices that analyze urine and stool for hydration, metabolism and disease markers. Priced between $200 and $400 with subscription fees, the gadgets aim to bring clinical‑grade monitoring into...
Replit CTO Explains Trusted Platform for Vibecoding Risks
Okay so we all now know vibecoding is a massive opportunity AND a massive risk. Now the question is what platform can you trust and why? Replit CTO lays it out ⬇️
AI Integrates Omics and Lifestyle to Forecast Health Outcomes
A multimodal AI agentic model that integrates electronic medical records, lifestyle, layers of biologic omics data to predict health outcomes and, with perturbations, "what if" scenarios a person improved lifestyle or took a medication @Cell_Metabolism https://t.co/qgVomjkipS https://t.co/JDaeYv0JGu
Interventional Radiologist Launches AI-Powered, IR-Specific Decision Support Platform
Interventional radiologist Dr. Syed Aziz Rahman unveiled VIRad.AI, an AI‑powered clinical decision‑support platform tailored for interventional radiology, at the Society of Interventional Radiology’s annual meeting. The tool combines an IR‑specific question bank, procedure reference library, device catalog, and an integrated...
LLMs Need a Full Operating System, Not QBasic
We really are still in the QBasic era of using LLMs. They really need a true Operating System. The beginnings of this is the LLM Harness, but there's so much more to be done.
MIT Study Reveals Vast Ecosystem of Niche AI Agents
A handful of agent platforms hog the headlines, but many function-specific agents are available to developers and users. MIT's latest study explores the broader agentic ecosystem. https://t.co/7bRibPFsRA via @ZDNET
AI Tool Predicts How New Drug Molecules Move Before Costly Lab Tests
University of Oregon researchers unveiled an AI‑driven simulation tool that predicts how novel drug molecules move and bind inside the body, using only their chemical structure. By integrating physics‑based energy data with machine‑learning sampling, the model delivers coarse‑grained motion pathways...
95% of Financial Firms Struggle to Build Effective AI
From FICO, not MIT: Nearly 95% of organizations are having a difficult time developing AI systems that serve the needs they were designed to address. The study is based on 254 C-suite leaders in financial enterprises around the world. via...
AI Can Enable Technofascism, Already Under Construction
AI does not inevitably lead to technofascism, "but it can enable it". #Palantir’s post "offers us a glimpse of the technofascist trajectory: not as a distant possibility, but as a world already under construction". Great piece by @MCoeckelbergh ⬇️ https://t.co/NZs6TkA5Gf
The Data Center Debate Cannot Hold
Maine Governor Janet Mills is weighing a historic bill that would temporarily ban the construction of new AI‑focused data centers across the state. The proposal stems from soaring electricity costs that many attribute to the power‑hungry servers powering generative AI...
Employees Left to Self‑teach AI Despite Employer Demands
Employers: You gotta know AI. Employees: Yes, help us learn more about AI. Employers: You're on your own. The majority of employees are learning through trial and error, without structure, benchmarks, or feedback loops, a new survey shows. via @forbes https://t.co/T4E7i6hsTB
GRPO: Reinforcement Fine‑Tuning for LLMs in 2026
How to fine-tune LLMs in 2026 https://t.co/JwehrYWi3o < looks at GRPO (Group Relative Policy Optimization), which is a form of reinforcement fine-tuning https://t.co/VNtbzppDy9
Dextall Unveils Kora Studio, AI Platform Cutting 315 Hours From Facade Design
Dextall opened early access to Kora Studio, an AI‑driven, Revit‑native facade design platform that eliminates an average of 315 hours of manual coordination on mid‑ and high‑rise projects. The tool integrates directly into Revit, automatically propagating design changes and generating...
OSCCA & DisplayWeek Host Applied Computer Vision Talks May 4
Our colleagues over at @opencvlive are holding their annual OSCCA conference in conjunction with @DisplayWeek in LA! Join @LearnOpenCV, @grbradsk, @DImagineering, @LKGGlass, @ultralytics and whole lot more on May 4th for a full day of applied computer vision talks. ⬇️⬇️⬇️...
Fine‑tuned Large Models Boost Robot Task Learning Efficiency
A new Science #Robotics study finds that “fine-tuned” large behavior models are more data-efficient and successful than single-task policies at teaching robots complex new tasks like coring and slicing an apple. @ToyotaResearch https://t.co/T6OugH8W0f https://t.co/m1iLz1mKf7