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
Guangxi Steel Plant Cuts Coal Use by 60,000 Tonnes Using AI Model
Guangxi Iron & Steel, a Liuzhou Steel Group subsidiary, has fully embedded the Xuantie AI model across its blast‑furnace complex, delivering an 8.5% rise in production efficiency, saving roughly 60,000 tonnes of standard coal and cutting 262,000 tonnes of CO₂ each year. The AI‑driven overhaul reshapes ore blending, molten‑iron transport and plate‑nesting, positioning the plant as a flagship of China’s smart‑manufacturing push.
Phenom Acquires Canadian Talent‑Assessment Platform Plum to Cut Bad Hires
Philadelphia‑based Phenom has completed its third acquisition of 2026, buying Canadian HR‑tech firm Plum. The deal, undisclosed financially, gives Phenom a full psychometric assessment stack to combat AI‑generated resumes and cut bad hires by up to one‑third.
CISA and Allies Publish First‑Ever Agentic AI Security Guidance for Critical Infrastructure
The U.S. Cybersecurity and Infrastructure Security Agency, together with the Australian Cyber Security Centre and other allies, issued a new guidance document on Friday outlining concrete security controls for the deployment of agentic AI in critical infrastructure. The guidance warns...
SAP Acquires Dremio and Prior Labs to Build an Open‑Source Lakehouse for Enterprise AI
SAP SE said it will acquire Dremio and Prior Labs, merging Dremio’s Iceberg‑native lakehouse with SAP Business Data Cloud and committing $1.17 bn to Prior Labs’ AI research. The move aims to eliminate data fragmentation that stalls enterprise AI projects.
Amazon Deploys Claude Code and OpenAI Codex to Entire Workforce
Amazon announced that Anthropic's Claude Code and OpenAI's Codex will be accessible to all corporate employees, with Claude Code live immediately and Codex slated for May 12. The rollout, managed through AWS Bedrock, ends a period of internal friction over...
Microsoft Patches VS Code to Stop Copilot Auto‑adding Co‑author Credit
Microsoft issued an emergency fix for VS Code that disables the default Copilot co‑author tag after developers reported the AI was being credited for code they wrote. The change, slated for the 1.119 release, reverses a March update that added the...
Singapore Boffins Get Diverse SIEMs Singing in Harmony with Agentic Rule Translation
Researchers from the National University of Singapore and Fudan University unveiled ARuleCon, an agentic retrieval‑augmented generation framework that translates security rules across five major SIEM platforms. The system pulls vendor documentation to resolve schema mismatches and runs Python‑based consistency checks...
IPhone 18 Pro Rumored to Pack Biggest Camera Overhaul Yet
Apple is rumored to unveil the iPhone 18 Pro with the most extensive camera hardware upgrades ever, including a variable‑aperture main lens and integrated Visual Intelligence. The changes could reshape how consumers choose premium smartphones and pressure rivals to accelerate...
Deepinfra Lands $107M in Funding to Build Out Its Dedicated Inference Cloud for Open-Source Models
Deepinfra Inc. announced a $107 million Series B round led by 500 Global, with participation from Nvidia, Samsung Next, and other investors. The funding will fuel the expansion of its purpose‑built inference cloud that runs open‑source AI models at scale. Deepinfra claims its...
How LLMs Actually Work
A new interactive guide walks readers through the end‑to‑end construction of large language models, from raw internet text to a conversational assistant, using Andrej Karpathy’s technical deep‑dive as its foundation. The guide emphasizes that each word a model emits is...
Citi Launches Arc AI Agent Platform to Automate Financial Services
Citi has unveiled Arc, a proprietary AI‑agent platform that lets developers build autonomous agents for multi‑step workflows across the bank. The system builds on a two‑year AI foundation and will initially be limited to internal developers before scaling to broader...
Accounting Firms Pour Millions Into AI to Transform Services
Leading accounting firms are allocating millions of dollars to artificial‑intelligence platforms, aiming to shift from compliance work to higher‑margin advisory services. The race pits large firms that buy ready‑made solutions against midsize players that are building proprietary tools, creating a...
Google AI Max Marks 1‑year Anniversary with AI Brief, Shopping & Travel Upgrades
Google celebrates the first anniversary of its AI Max advertising tool by launching AI Brief, extending the platform to Shopping and travel campaigns, and adding mandatory disclaimer support for final URL expansion. The upgrades aim to give advertisers tighter brand...
Deviation Capital Spins Out From Two Sigma with $2 B AUM to Target Data‑Driven Startups
Deviation Capital announced its launch as an independent venture firm after spinning out of Two Sigma Ventures, taking control of roughly $2 billion in assets under management. Led by former TSV partners, the firm will focus on early‑stage companies that leverage...
Finom Unveils Standalone AI Accounting Tool in Germany, Pricing at $15/Month
Finom has released a standalone version of its AI‑driven accounting platform in Germany, allowing freelancers and small businesses to connect any bank via PSD2. The service is free for six months, then costs €14 ($15) per month, positioning it 36%‑76%...
Anthropic, Blackstone, Hellman & Friedman and Goldman Sachs Launch AI Services Firm to Deploy Claude
Anthropic announced a joint venture with Blackstone, Hellman & Friedman and Goldman Sachs to create an AI‑native enterprise services firm. Backed by a consortium of alternative asset managers, the new company will embed Claude into mid‑size businesses, addressing a surge in demand for...
Google’s Pentagon AI Deal Triggers Employee Backlash Over Military Use
Google signed a Pentagon contract that permits its AI models to be used in classified military systems for any lawful purpose, prompting a surge of internal dissent. Around a thousand employees have signed an open letter demanding the company halt...
Google On Keyword Fragmentation And User Needs In AI Search via @Sejournal, @Martinibuster
Google’s search product lead Liz Reid told Bloomberg’s Odd Lots that AI Mode and AI Overviews are prompting users to type longer, need‑based queries. These queries are fragmented by Google into multiple precise sub‑queries, which are then answered with synthesized...
Pentagon Seeks Smarter, Self-Organizing Drones as Autonomous-Warfare Budget Is Poised to Skyrocket
The Pentagon plans to increase its autonomous‑warfare budget to $54 billion for FY 2027, up from $226 million this year, aiming to field swarms of AI‑driven drones. DARPA’s new Requests for Information target self‑organizing robots that can think locally and collaborate without central...
CARSOME and Darwinbox Partner to Reinvent Employee Experience with AI-First HR and Native Payroll for 3,000 Employees Across Southeast Asia
CARSOME, Southeast Asia’s largest integrated car‑e‑commerce platform, has appointed Darwinbox as its strategic HR‑technology partner to deploy an AI‑first Human Capital Management suite for over 3,000 employees across Malaysia, Singapore, Indonesia and Thailand. The rollout began with core HCM functions...
Join Us at Big Technology's AI Summit in San Francisco on 6-18
Big Technology announced its inaugural AI Summit, scheduled for June 18, 2026 at San Francisco’s historic Commonwealth Club. The event aims to bring together AI innovators, investors, and enterprise leaders for a day of keynotes, panels, and networking. Confirmed speakers...
ThredUp AI Agents Personalize Shopping in Real Time
ThredUp introduced its first agentic AI experience, assigning real‑time personal agents to shoppers on its resale platform. The agents ingest clickstream data across channels and use reinforcement learning to reshape product listings instantly, aiming to boost conversion and retention. The...
Open‑source Models Behind Firewalls Could End Corporate Opacity
I actually think that some things you want to keep illegible *from other organizations* but IMO there is an opportunity to do this behind a firewall with open source models. To selectively send context out only when needed. I'd...
How Is Agentic AI Revolutionizing Worker Safety in the Field?
Agentic AI is reshaping worker safety in the energy sector by deploying autonomous, collaborative agents that create dynamic safety perimeters, predict hazards, and coordinate real‑time emergency responses. Multi‑agent systems combine specialized AI nodes—environmental monitoring, equipment status, compliance, and emergency response—under...
Newly Formed Tripartite Jobs Council to Support Workers and Businesses Through AI Transformation
Singapore’s Ministry of Manpower, NTUC and the Singapore National Employers Federation are launching a Tripartite Jobs Council (TJC) to guide the nation’s AI transition. The council will coordinate AI‑responsible adoption, sector‑wide training and targeted support for workers whose roles are...
FIS and Anthropic Collaborate to Enable Agent-First Banks
FIS Global and AI firm Anthropic have co‑developed a Financial Crimes AI Agent that automates anti‑money‑laundering investigations, delivering results in minutes. The agent, built with Anthropic’s Claude models and FIS’s unified data infrastructure, will be first deployed at BMO and...
FIS Global Rolls Out AI‑Powered Cloud Platform for 68 Million 401(k) Participants
FIS Global Retirement announced a July launch of an AI‑driven cloud record‑keeping platform that merges its Omni and Relius systems. The new solution will serve 68 million 401(k) participants across 650 clients, promising faster integration of third‑party apps and a modernized...
Hirschbach to Deploy 500 Aurora‑Powered Autonomous Trucks by 2027
Hirschbach Motor Lines announced a memorandum of understanding to acquire 500 Aurora‑equipped autonomous trucks, with deliveries slated for 2027. The deal promises up to 500 million driverless miles and a revenue stream for Aurora worth hundreds of millions of dollars, marking...
Anthropic Secures $1.5B Backing for New Enterprise AI Venture Targeting Sales Teams
Anthropic announced a new venture to sell its Claude generative‑AI model to enterprise customers, backed by Goldman Sachs, Blackstone, Hellman & Friedman and a slate of alternative asset managers. A Wall Street Journal report estimates the partners will commit about...
DMALL Rolls Out AI-Driven Platform to 87 Cold Storage Stores in Southeast Asia
DMALL Inc. completed the integration of 87 Cold Storage Singapore supermarkets, hypermarkets and express stores into its unified AI-driven retail operating system within seven months. The rollout, the latest in DMALL’s Southeast Asian push, promises tighter supply‑chain visibility and faster...
Databricks Ramps up A/NZ Investment with $420M Pledge
Databricks announced a US$300 million (A$420 million) investment in Australia and New Zealand over three years, funding a new 22,000‑sq‑ft regional headquarters in Sydney and a large upskilling program. The company aims to train 100,000 people in AI and data analytics through bootcamps...
Siemens Commits $1 B to U.S. Manufacturing, Adding 2,200 Jobs by 2028
Siemens said it has reached $1 billion in U.S. manufacturing investments that will roll out in 2026, generate more than 2,200 jobs by 2028 and deepen its digital‑automation footprint across 25 sites. The move underscores the firm’s push to strengthen American...
Ncontracts Launches Nquiry AI Platform for Instant, Auditable Compliance Answers
Ncontracts introduced Nquiry, an AI‑powered regulatory intelligence platform, on May 4, 2026 in Nashville. The tool promises to return cited, auditable answers to complex compliance questions within minutes, drawing on 17 years of proprietary data. Legal and risk teams in...
Tanzania Nears Launch of Swahili LLM to Power Government Services
Minister Angellah Kairuki said Tanzania is close to releasing a Swahili large‑language model that will enable AI‑driven public services in fluent Kiswahili. The rollout is part of a broader push for digital inclusion, cybersecurity and data‑driven growth in the communications...
AI Adoption Linked to Mass Layoffs and Longer Work Hours, New Data Shows
Atlassian, Block and Snap cited AI as a factor in recent mass layoffs, while Sharebite data reveals a surge in after‑hours corporate food orders, suggesting AI is extending, not shrinking, workdays. The findings revive the Pentagon Pizza theory and raise...
Tesla Already Has 10 Billion Miles of Human Driving Data for FSD.
I feel like the writers at The Verge get stupider every week. What Elon said, in response to a thread about Nvidia Alpamayo, was that safe unsupervised self-driving required a training set somewhere on the order of 10 billion miles....
Anthropic's Edge: AI-Driven Workflow Overhaul Fuels Success
I was pretty sure Anthropic uses Mythos internally, and Boris confirmed that. They also use Opus 4.7 today to write all their code, all their sql... Anthropic's advantage today is that they've reorganized the processes and workflows around AI. If you look...
Why Public Sector AI Uptake Keeps Stalling
Australian government agencies are eager to experiment with AI, launching numerous pilots across predictive analytics and natural‑language tools. However, few projects move beyond proof‑of‑concept to production, reflecting a global pattern amplified by Australia’s fiscal pressures, complex procurement and data‑sovereignty expectations....
Organizational AI Remains Jagged, Lacking Clear Terminology
For individual AI use, the jagged frontier is increasingly well understood. In multi-agent workflows in organizations, AI is jagged in ways that have not been well identified yet. In fact, we don't even have a vocabulary around multi-agent systems & the...
Voice AI Is Reshaping How We Interact with Technology
pretty excited for voice models to get great its interesting to watch how people are already starting to change the way they interface with AI
Pentagon Signs Contracts with Seven AI Firms to Bring Tools Onto Classified Networks
The Pentagon finalized agreements with seven artificial‑intelligence companies, allowing their products to run on secret and top‑secret networks. The move expands the AI supplier pool, excludes Anthropic amid a supply‑chain dispute, and speeds approvals from 18 months to under three...
Dubai Unveils Agentic AI Incubators, Funding Future Economy
Dubai is launching dedicated incubators and funds specifically for Agentic AI startups to build the economy of tomorrow. A massive leap forward for global AI adoption. 🇦🇪 This is going to be a game-changer for the global tech ecosystem. 🌍💡 ...
Kids Can Create Shareable AI Agents and Earn Online
I want to get my 8 year old to start building stuff with agents that she actually can share with her class and teachers. Any ideas? Maybe she can make her first dollar online too.
Ouster Unveils Rev8 OS, First Native‑Color Lidar for Autonomous Systems
Ouster Inc. introduced the Rev8 OS sensor family, its first native‑color lidar that doubles the range and resolution of its prior generation. The launch follows a $38 million StereoLabs acquisition and promises to accelerate commercial autonomy across vehicles and robotics.
Workhuman Debuts AI‑Powered ‘Future Leaders’ Tool to Spot High‑Potential Talent
Workhuman introduced Future Leaders, an AI‑powered platform that leverages its Ascend AI engine to surface high‑potential employees well before they enter formal leadership pipelines. The solution promises to cut the cost of senior‑level mis‑hires by using real‑time workplace interaction data.
Suno AI Moves to Overhaul Songkick Concert Discovery with AI Integration
Suno AI, the AI music startup that recently closed a $250 million Series C round, is set to reimagine Warner Music Group’s former concert‑discovery platform Songkick with artificial intelligence. The plan hinges on leveraging Songkick’s user data and Suno’s own 2 million paid...
ComfyUI Secures $30 Million, Valuation Hits $500 Million
ComfyUI, a startup that gives creators granular control over AI‑generated media, closed a $30 million financing round that values the company at $500 million. The round was led by Craft Ventures with participation from Pace Capital, Chemistry and TruArrow. The capital will...
Samsung Shifts Focus to AI‑Driven Data‑Center Memory and Storage
Samsung Electronics announced it will prioritize memory and storage production for AI‑focused data centers, marking a strategic pivot from consumer devices. The move reflects soaring AI demand and positions Samsung as a key supplier for big‑data workloads.
IAB Europe’s 2026 Programmatic Day Highlights CTV Ownership Gaps and AI‑Driven ‘Messy Middle’
IAB Europe convened its Virtual Programmatic Day 2026 at Google’s London office, drawing hundreds of in‑person attendees and a larger online audience. The two‑panel event dissected fragmented CTV budget ownership and the rise of AI in the programmatic ‘messy middle’,...
Google Rolls Out Gemini Embedding 2, a Unified Multimodal Vector Model
Google has made Gemini Embedding 2 generally available via its Gemini API and Enterprise Agent Platform. The model maps text, images, video, audio and documents into a single semantic space, supporting over 100 languages and delivering up to a 27‑point lift...