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
Nestlé Trials AI Technology to Tackle Factory Food Waste
Nestlé has piloted Zest’s AI platform across its factories, instantly linking fragmented production data to pinpoint and reduce food waste. The system uncovered four times more edible surplus than manual checks, redirecting 201.9 tonnes—equivalent to about 480,000 meals—to people in need and generating roughly $1.3 million in additional revenue. Over 200 tonnes of surplus were saved, and the pilot cut waste‑identification time by half. The technology is slated for broader rollout across Nestlé’s product lines, from confectionery to coffee.
AI Banking Success Requires Region‑Specific, Production‑Scale Deployments
AI is not “one-size-fits-all,” and we need to recognize that use cases that work in SEA won’t work in LatAm. While the AI spend in banking is projected to surge over 10x from $35 billion in 2023 to $368 billion by...
Njordium AI Blocks Fake Invoices and Fraudulent Payments
Njordium Cyber Group unveiled an AI‑driven Fraud Detection Module integrated into its new Vendor Management System, instantly flagging fake invoices, phantom services and abnormal pricing. The self‑learning engine extracts data from PDFs, OCR scans, Excel, XML and email, routing high‑confidence...
Synopsys Supports Arm AGI CPU Development
Synopsys announced a deep collaboration with Arm to support the development of the new Arm AGI CPU, built on the Neoverse CSS V3 architecture. The partnership leverages Synopsys’s full‑stack design portfolio—including VCS simulation, Fusion Compiler, IC Validator, PrimeTime, RedHawk‑SC, silicon‑proven...
The Sequence AI of the Week #830: The Quiet Ambush: Inside the Amazing MiMo-V2-Pro Aka Hunter Alpha
The blog spotlights MiMo‑V2‑Pro, also known as Hunter Alpha, a stealthy AI agent that appeared as an unnamed OpenRouter API endpoint. Unlike traditional chat‑focused models, MiMo‑V2‑Pro showcases hybrid attention and multi‑token prediction, positioning it for autonomous, tool‑driven workflows. The author...
AI Fails when Built on Non‑AI Infrastructure
Most enterprise AI initiatives don’t fail because of bad models. They fail because they’re running on infrastructure that was never designed for AI. A traditional data center is not the same thing as AI infrastructure. This came up directly in my conversation with...
Notch Raises $30M to Expand AI Operating System for Regulated Industries
Notch Ltd., a former specialty insurer turned AI startup, announced a $30 million Series A round to accelerate its AI operating system for regulated sectors, especially insurance, across the United States. The platform provides transparent, auditable AI agents that automate both front‑office...
The Secretary of War Didn’t Really Mean It, Contends US Government Lawyer as Anthropic Gets Its First Day in Court...
Anthropic appeared before U.S. District Judge Rita Lin to challenge a Trump 2.0‑issued blanket ban that bars any contractor working with the Department of War from using its AI technology. The judge highlighted that Secretary of War Pete Hegseth’s social‑media announcement...
Duos Edge AI and Seimitsu Partner to Strengthen Digital Infrastructure Across Georgia
Duos Technologies Group, through its Duos Edge AI subsidiary, announced a strategic partnership with fiber provider Seimitsu to build edge computing infrastructure across Georgia. The collaboration will embed modular Edge AI nodes onto Seimitsu’s 25‑terabit low‑latency fiber network, delivering ultra‑low...
Aviva Pilots ChatGPT App for Home Insurance Quotes
Aviva has launched a pilot ChatGPT app that delivers instant home‑insurance quotes within minutes. The app, available through the ChatGPT App Store, asks users basic details—name, address, contact and policy type—to generate a quote for Aviva’s Signature Home Insurance product....
PossibleNOW and Convoso Announce Strategic Partnership to Align Compliance and AI-Powered Outbound Performance
PossibleNOW, a compliance‑focused SaaS provider, has entered a strategic partnership with Convoso, an AI‑driven outbound contact‑center platform. The alliance merges PossibleNOW’s Do‑Not‑Contact (DNC) governance tools with Convoso’s predictive dialing and workflow automation, allowing customers to scale outreach while mitigating TCPA...
WuXi Biologics Reports Record 2025 Annual Results, Operational Excellence Driven by Digital-Native Architecture
WuXi Biologics reported record 2025 results, with revenue up 16.7% and IFRS gross profit climbing 30.9%, lifting its gross margin to 46%. The growth was driven by expanding research, development and manufacturing contracts and tighter cost control across its global...
Approaching.ai Brings in Top Scientists to Capture AI’s Inference Boom
Approaching.ai announced the appointment of Academician Wei‑Min Zheng as Chief Scientific Advisor and Professor Yongwei Wu as Chief Scientist, bolstering its technical leadership in AI inference. Both experts bring decades of high‑performance computing and distributed systems experience, reinforcing the company’s...
Nevermined Unveils Card Delegation for AI Agents, Enabling Autonomous Payments
Nevermined introduced persistent card delegation, letting AI agents hold delegated credit cards and make real‑world payments without human approval. The feature embeds spending limits, duration controls and revocability directly into the infrastructure, mirroring corporate expense policies. By enabling continuous, automated...
Glovo CTO Shiro Theuri Is a Guardian for Benefiting From AI
Glovo’s chief technology officer, Shiro Theuri, is steering the Barcelona‑based delivery platform through a rapid AI integration across its engineering teams. She reports high AI penetration, leading to faster, higher‑quality code while instituting responsible‑AI guardrails and human‑in‑the‑loop reviews. Theuri also...
AI Silently Patched Critical Bug While I Slept
Claude went into my computer at 1:15am and changed the code in something we’re building. I was asleep I never told it to do it. It just went in on its own and fixed an issue. In the morning when I realised what...
Advanced Deep Learning Interview Questions #4 - The I/O Starvation Trap
During a senior ML engineer interview at Meta, candidates are asked why training speed stalls after moving deep‑learning workloads to a large AWS GPU cluster. Although the expensive GPU instances launch correctly, the iteration rate does not improve. The hidden...
The Struggle to Prove AI Productivity Gains
Enterprises are grappling with how to quantify AI‑driven gains in software engineering, even as boardrooms shift from mere tool adoption to demonstrable output. A Multitudes survey of 700+ engineers found 75% struggle to measure AI impact, and only 31% have...
First Innovate for Ireland National Centre Launched – ‘Decarb-AI’
Today Ireland unveiled its first Innovate for Ireland national centre, Decarb‑AI, a €5.7 million (≈$6.2 million) initiative partnered with AIB and Research Ireland to accelerate AI‑driven climate resilience. The centre will host 30 international iScholars across three intakes, each pursuing fully funded...
Can You Trust the Spec? The Risky Future of Agent-Compiled Software
OpenAI’s Symphony orchestrator lets developers describe software in a natural‑language specification and have AI agents compile it on demand, bypassing traditional installers. The approach echoes StrongDM Attractor’s spec‑driven workflow and promises on‑the‑fly, customized builds for each user. Critics warn that...
Basecamp Research Unveils Trillion Gene Atlas to Boost Longevity Drug Discovery
Basecamp Research has launched the Trillion Gene Atlas, a platform that will collect and model genetic information from more than 100 million species, expanding known evolutionary diversity by roughly 100‑fold. The initiative aims to give AI models a vastly broader biological...
Hachette Pulls Horror Novel "Shy Girl" Over AI‑Generated Text Claims
Hachette Book Group has cancelled the US publication and discontinued the UK edition of Mia Ballard’s horror novel "Shy Girl" after AI‑detection tools flagged the majority of its prose as machine‑generated. The move marks the first known instance of a major...
TSA Rolls Out Touchless PreCheck ID at 65 Airports, Starting with Milwaukee
The Transportation Security Administration has begun deploying its Touchless PreCheck ID service at 65 airports, with Milwaukee Mitchell International Airport as the first site. The facial‑recognition system lets eligible PreCheck members verify identity without a physical ID, a move designed...
Arm Launches First Production‑ready AGI CPU for Next‑gen AI Infrastructure
Arm introduced its first production‑ready AGI CPU, positioning the chip as a fresh platform for building scalable AI services. The company did not disclose performance metrics, pricing or rollout timeline, leaving CTOs to await further specifics.
Consultants Grapple with Scaling Dependable AI for Fortune‑50 Firms
Anil Pantangi, a senior partner at a leading management‑consulting firm, outlined how consulting teams are tackling the architecture, data‑governance and talent challenges of deploying dependable AI across Fortune‑50 enterprises. He stressed that the biggest friction points are legacy systems, risk‑averse...
YY Group Appoints Arros AI Co‑founder Kai Yang as Chief AI Scientist
YY Group Holding Limited announced the appointment of Arros AI co‑founder Kai Yang as its first Chief AI Scientist, effective April 1, 2026. The hire is tied to the company’s push to embed AI‑driven recruiting tools into its YY Circle platform as...
6 Types of AI Content Moderation and How They Work
AI-powered content moderation is reshaping how platforms manage user‑generated material, moving from manual review to a blend of automated and human processes. Six moderation models—pre‑moderation, post‑moderation, reactive, distributed, user‑only, and hybrid—define where and how AI intervenes, from blocking offensive posts...
No Source Confirms Google AI Data Center Plan Near Oklahoma City, Recall Talk Unverified
A review of the eight supplied sources found no verifiable information about a Google‑planned AI data center on 827 acres near Oklahoma City or any related council recall effort. The absence of confirmed details underscores the difficulty of tracking tech‑driven...
Signus.ai Launches AI-Native Contract Platform, Partners with AAA
Signus.ai announced the public launch of its AI-native contract intelligence and e‑signature platform and a collaboration with the American Arbitration Association. The partnership embeds AAA’s ClauseBuilder® AI into contract workflows, aiming to tighten dispute‑resolution clauses and reduce post‑signing risk.
Workday Unveils Sana AI Engine to Automate HR and Finance Workflows
Workday introduced Sana, a unified AI interface that embeds a conversational layer into its HR and finance suite, enabling users to execute tasks across applications. The launch includes a self‑service agent with more than 300 pre‑built skills and a shift...
Arm Debuts AI‑Focused AGI CPU, Locks In Meta, OpenAI, Cloudflare as Early Users
Arm Holdings introduced the Arm AGI CPU, its first in‑house silicon designed for large‑scale AI workloads, claiming more than twice the performance‑per‑watt of comparable x86 racks. Meta, OpenAI, Cloudflare and other tech leaders signed on as early adopters, signaling a...
Cursor's $29B Valuation Tested as Investors and Demand Falter
Cursor, the AI‑driven code generation platform valued at $29 billion, is seeing investor enthusiasm cool and market demand soften. The shift highlights broader uncertainty for AI‑assisted development tools that power DevOps pipelines.
Xcel Energy, Google Team up on 1.6 GW Long‑duration Storage Project Amid AI‑driven Power Surge
Xcel Energy and Google have signed a deal to build a 1.6 GW multi‑technology generation stack in Pine Island, Minnesota, featuring iron‑air batteries for long‑duration energy storage. The partnership reflects soaring AI compute demand that is pushing utilities to seek multi‑day...
Zoox Expands Robotaxi Service to Austin and Miami, Boosting U.S. Footprint
Amazon-owned Zoox announced the rollout of its purpose‑built robotaxi service for employees, families and friends in Austin, Texas and Miami, Florida later this year. The move expands Zoox’s early‑rider program to four U.S. cities and follows a cumulative 1.98 million autonomous...
Alibaba Q4 2025 Earnings Show Two‑Thirds Profit Drop as Cloud, AI Drive Growth
Alibaba reported a roughly two‑thirds year‑over‑year plunge in net income for Q4 2025, with e‑commerce revenue growth slowing to 6% and cloud revenue jumping 36%. The group said the loss reflects heavy investment in cloud, AI and quick‑commerce, positioning the...
Disney CEO Josh D’Amaro Confronts Early Setbacks as OpenAI Deal Collapses
Disney’s newly appointed chief executive, Josh D’Amaro, is grappling with two high‑profile setbacks in his first week: OpenAI abruptly terminated its Sora video‑AI tool, ending a $1 billion agreement with Disney, and a planned Fortnite collaboration with Epic Games has stalled...
TurboQuant Compresses AI 6×, Speeds 8×, Runs Locally
Absolute GAME CHANGER for local AI 🤯 @GoogleDeepMind’s new TurboQuant algorithm achieves extreme AI compression, completely redefining model efficiency. Usually, shrinking an AI makes it “dumber.” Not this time. It’s an algorithm that makes LLMs smaller and faster without losing quality. The benchmarks: → Achieves a...
OpenAI Pulls the Plug on Sora, Ending $1 Billion Disney Pact
OpenAI announced today that it is shutting down Sora, its AI video‑generation product, and terminating a $1 billion partnership with Disney. The move reflects mounting compute expenses and a strategic pivot toward enterprise and developer‑focused AI offerings.
AI Agents Show Progress Yet Reliability Gaps Stall Data‑Driven Rollouts
In the last 24 hours, industry analysts noted that while autonomous AI agents are gaining capabilities, persistent reliability issues are limiting their adoption in data‑intensive environments. The gap between performance and trust is prompting firms to pause large‑scale rollouts.
Kleiner Perkins Raises $3.5 B Across Two Funds to Double Down on AI
Kleiner Perkins announced a $3.5 billion raise – $1 billion for its 22nd early‑stage fund and $2.5 billion for a late‑stage vehicle – to accelerate AI‑focused investments. The haul, the firm’s largest since 2024, comes as venture capital pivots toward generative‑AI opportunities.
ServiceNow Shares Fall 4.4% as Anthropic Unveils Remote‑Control Claude Agent
ServiceNow stock fell 4.4% after Anthropic announced a new Claude AI feature that lets the chatbot operate a computer remotely. The move triggered a 4% drop in the IGV SaaS ETF and revived worries that generative‑AI agents could erode the...
AI Agents Become Profit Drivers, Demand New Accountability
Agentic AI is starting to change something fundamental in financial services: who actually drives profit and loss. We are moving beyond models that inform decisions. AI agents are beginning to price risk, approve loans, detect fraud, handle claims, and manage...
OpenAI Announces Closure of Sora App Amid Unclear Details
OpenAI has announced that it will shut down its Sora app, though the company provided no details on the timeline, user migration plan, or financial impact. The lack of information leaves analysts and developers questioning the strategic rationale behind the...
M&S Rolls Out AI Tools to 11,000 Colleagues Including All Store Managers
Marks & Spencer is rolling out Microsoft 365 Copilot to 11,000 employees, including every store manager, as part of a broader digital transformation. The AI suite will automate routine tasks such as sales insight compilation, meeting‑note summarisation, shift‑roster creation and...
Shane Heraty to Spearhead Cisco’s AI Infrastructure Push in MEA–TRC
Cisco has named Shane Heraty as Vice President and General Manager for its Middle East, Africa, Türkiye, Romania and CIS region, effective May 1. Heraty will drive Cisco’s AI infrastructure, security and workplace modernization strategy across a market of over 1.5 billion...
Elemica Expands AI-Driven Capabilities to Transform Customer Onboarding and Order Orchestration
Elemica announced expanded AI-driven capabilities across its supply‑chain execution platform, adding agentic AI to its Launchpad onboarding service and new self‑service onboarding tools. The technology enables companies to onboard customers and order formats without heavy IT or data‑science involvement, reducing...
Akamai Launches AI Grid Intelligent Orchestration
Akamai announced the first global‑scale implementation of Nvidia’s AI Grid, deploying thousands of RTX PRO 6000 Blackwell Server Edition GPUs across its 4,400‑node edge network. The AI Grid adds an intelligent orchestrator that routes inference requests to the optimal compute tier, slashing...
AI‑Generated Grok Ad Wins Elon Musk’s $1M Contest
This #video won Elon Musk's contest for the best AI-generated #Grok ad. The #Grokker who prompted it won $1 million. #artificialintelligence #AI #Aiartowork #AIart️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️️ #AIvideo #aiad @lexfridman @KirkDBorne @Ronald_vanLoon @erikbryn @antgrasso @sallyeaves @Nicochan33 @HaroldSinnott @mvollmer1 @marcusborba @CatherineAdenle @vinod1975 @YuHelenYu @ingliguori @fogoros @ylecun...
Claude AI Gains Always‑On Background Agent Capabilities
@AnthropicAI is pushing @claudeai toward an “always-on” agent — it can run tasks in the background, use your computer, and be controlled remotely. It’s taken a cue from @openclaw, for sure, but with tighter controls around access. 🔗➤ https://t.co/RCTn7FiqvA
AI Amplifies Auditors, Enabling Full Transaction Coverage
EY and PwC aren't replacing auditors with AI. They're giving auditors 100% transaction coverage instead of 5% sampling. AI catches the anomaly. The auditor interprets the meaning. That's not disruption. That's amplification. https://t.co/5bdaBFwzae