Today's AI Pulse
-1.png)
Anthropic shuts down Fable 5 and Mythos 5 models under U.S. export controls
Anthropic abruptly disabled its flagship Fable 5 and Mythos 5 models to comply with a U.S. export‑control directive, underscoring the vulnerability of firms that depend on closed‑source AI services. The move triggered a sharp rally in Chinese open‑source offerings as investors shifted to downloadable models that can run on‑premise.
Also developing:
By the numbers: Respond raises $62.5M Series B
"Hey Grok, Find a Charger and Tell Me the Plymouth Score": Tesla AI Bot Changes Driving for Good
Tesla has rolled out Grok, a large‑language‑model voice assistant built into the Model Y. By long‑pressing the steering‑wheel mic—or simply saying “Hey, Grok”—drivers can ask complex, multi‑step requests such as routing to a destination, scheduling charging stops and even checking sports scores. The system taps the car’s manual, real‑time telemetry and internet sources to execute commands or guide users to the appropriate touchscreen menu. Early testing shows Grok handling most requests reliably, with only occasional missteps on niche queries.
A Practical Roadmap to Agentic Programming
Agentic Programming: A Roadmap by Shittu Olumide @TeachTheMachine Learn more: https://t.co/CrubN8o1xG #LLM #GenerativeAI #ArtificialIntelligence #MachineLearning https://t.co/xA77KORwdW
UK Military Looks at Allowing Lethal Strikes without Human Approval
The UK Ministry of Defence is reviewing its rules to allow lethal autonomous strikes without a human operator’s final approval. The policy shift would relax the current human‑in‑the‑loop requirement for certain weapon systems, including drone swarms and loitering munitions. Officials...
Amidst Codex's Strong Performance... Anthropic Releases Claude Opus 4.8 Just 42 Days Later
Anthropic unveiled Claude Opus 4.8 on May 28, just 42 days after its 4.7 predecessor, marking the fastest model update in the company’s history. The new version keeps the same API pricing but promises modest yet clear gains, especially in code honesty and...
AMD to Invest $10 Billion+ in Taiwan
AMD announced a commitment of over $10 billion to expand its Taiwan ecosystem, targeting advanced packaging and AI‑focused silicon manufacturing. The funding will accelerate development of Elevated Fanout Bridge (EFB) interconnects and support the Helios rack‑scale AI platform slated for the...
Arteris Technology Adopted by Li Auto for Intelligent Vehicles
Li Auto has integrated Arteris’s FlexNoC 5 network‑on‑chip IP and Magillem integration automation into the proprietary SoC powering its new L9 Livis SUV. The SoC delivers 2,560 TOPS of AI compute, supporting autonomous driving, ADAS, and immersive in‑vehicle experiences. Arteris’s technology enables efficient...
Pennsylvania Gov. Shapiro Mandates BYOE for AI Data Centers, Sets Clean Energy Targets
Pennsylvania Governor Josh Shapiro announced a "bring your own energy" (BYOE) rule that forces high‑power AI data centers to finance and source their own electricity, with clean‑firm targets of 10% by 2027, 14.5% by 2030 and 32% by 2035. The...
Anthropic Hits $965B Valuation; AI Complexity Plagues Indian Companies
Anthropic announced a $65 billion Series H round that lifted its post‑money valuation to $965 billion, briefly overtaking OpenAI in the AI frontier‑model race. The round was led by Altimeter Capital, Dragoneer, Greenoaks and Sequoia and includes $5 billion from Amazon. A Freshworks study...
AI Tool of the Week: This OpenAI Tool Tells You if an Image Is AI-Generated
OpenAI has launched a free web tool called OpenAI Verify that lets users determine whether an image was generated by artificial intelligence. The service scans uploaded files for two provenance signals: C2PA Content Credentials, a cryptographic metadata layer, and SynthID,...
The AI Antidote
Executives are urged to shift AI focus from flashy features to measurable outcomes. Companies should answer whether customers truly demand AI and how the technology differentiates them, otherwise spend becomes a costly experiment. A three‑bucket framework—internal productivity, engineering efficiency, and...
From Resilience to Survivability: How AI Forces a Rethink of Business Continuity
Artificial intelligence is reshaping business continuity, pushing firms from traditional resilience to "operational survivability." Equinix’s recent report warns that Global 2000 companies lose roughly $400 billion in downtime each year—about $540,000 per hour—and argues that redundancy alone won’t suffice as AI...
Google Leans on Full-Stack AI Edge, Says Cloud Customers Could Save $1B+ a Year by Shifting 80% of Workloads to...
Google is promoting its full‑stack AI advantage, saying cloud customers could save more than $1 billion annually by moving 80% of AI workloads to Gemini 3.5 Flash and other frontier models. Monthly usage of Google’s AI services has jumped sevenfold to 3.2 quadrillion tokens...
Google’s Gemini 3.5 Flash Faces Scrutiny Over Hallucinations and Search Accuracy
Google’s newest Gemini 3.5 Flash model, rolled out at I/O, is being challenged for its hallucination rate and basic spelling mistakes. Internal benchmarks show accuracy between 68.8% and 83.8%, prompting publishers and regulators to question the reliability of AI‑generated search...
Arlo Technologies Trades at Half Its Fair‑Value as AI‑Driven Services Replace Camera Sales
Arlo Technologies' stock is trading around $13.50, roughly half of analysts' $24 fair‑value estimate, while the company pivots from selling cameras to licensing AI‑driven security software. The shift, highlighted by a service‑only deal with Samsung and 28% ARR growth, has...
Snowflake Posts Blowout Quarter and Seals $6 B AWS Deal, CFO‑Driven Capital Strategy in Spotlight
Snowflake delivered a record first‑quarter with product revenue of $1.33 billion, up 34% YoY, and announced a five‑year $6 billion agreement with Amazon Web Services. CFO Brian Robins emphasized a disciplined capital return plan and a modest margin headwind from recent acquisitions.
Kakunin Launches EU AI Act‑Compliant Identity Platform for Autonomous Agents
Kakunin has rolled out a cryptographic identity platform that issues X.509 certificates to autonomous AI agents, aligning with the EU AI Act and MiCA regulations. The service bundles identity, event monitoring and audit‑ready reporting, aiming to cut weeks off compliance...
How AI Can Lead To False Arrests & Wrongful Convictions
AI‑driven surveillance and facial‑recognition systems have produced stark miscarriages of justice, exemplified by a Maryland teen whose chip bag was flagged as a gun and a Tennessee grandmother wrongfully jailed for five months after a facial‑recognition error. These tools generate...
AWS Reportedly to Tuck Elon Musk's Grok Into Bedrock, Despite Zero Enterprise Demand
AWS is reportedly preparing to list Elon Musk’s Grok LLM on its Bedrock marketplace despite a clear lack of enterprise interest. Interviews with cloud‑contract signers reveal banks and large firms actively reject Grok, citing performance and reputational concerns. The move...
HIMSSCast: Leaders Beyond CFOs Are Making Investment Decisions in AI
Qventus released its second annual CIO report, revealing that 45% of chief information officers are now the primary decision‑makers for AI purchases in health systems. The report emphasizes treating AI as core infrastructure rather than a standalone innovation to unlock...
Judgment Is the Currency of a Good Lawyer. The Work Now Is to Define It
The article argues that AI is rapidly automating the visible tasks of commercial lawyers—research, drafting, and redlining—while exposing the invisible, judgment‑driven work that actually closes deals. It distinguishes crystallized intelligence (knowledge) from fluid intelligence (real‑time decision‑making), noting that AI cheapens...
AI Tackles the Paperwork Problem Blocking America’s Housing Permits
Denver approved a five‑year, $4.6 million contract with CivCheck, an AI‑enabled plan‑review platform, to automate the early screening of building‑permit applications. The tool flags missing documents and incomplete fields before plans reach human reviewers, aiming to lift first‑try approval rates from...
Okta Writes Its Own License to Kill Rogue AI Agents
Okta announced a new identity‑centric control layer for autonomous AI agents, positioning itself as the "kill switch" for rogue bots. The company highlighted that 92% of executives report moderate or widespread AI‑agent use, yet only 22% have tied those agents...
AI Services Sacrifice Five‑nine Uptime for Double Capacity
Great Stanford CS153 course... Amin Vahdat who runs Google's AI infrastructure and the TPU program behind Gemini. His point on reliability: "Five nines (99.999%) = 30 sec of downtime/year - but it costs you 2N redundancy, so half your power sits idle. Three...
Zhuoyu Technology European Headquarters Officially Opens, Announces New Cooperation with Volkswagen
Zhuoyu Technology inaugurated its European headquarters in Munich on May 27, unveiling a native multimodal foundation model designed for intelligent driving across passenger cars, heavy trucks, buses, and unmanned logistics. The company announced a proof‑of‑concept partnership with Volkswagen Group, integrating its...
Shoppable AI Data Sets New Standard For Media Buyers
Shopsense AI is preparing to launch its Shoppable Intelligence Model (SIM), a multimodal AI engine that links audience insights directly to commerce outcomes. In early testing, SIM outperformed OpenAI's CLIP and Google's SigLIP2 by 25‑50% on product‑discovery benchmarks and delivered...
Zocks, Jump Expand Advisor Reach with New Enterprise Integrations
Zocks has secured an exclusive AI‑assistant partnership with Hightower, covering its 650‑advisor network that manages roughly $352 billion. Jump was chosen as the AI operating system for Equitable Advisors, embedding intelligence across the firm’s client‑workflow stack. Both firms tout more than...
AI Gives Answers; Koji Builds True Problem Solvers
AI isn't making kids smarter. It's making it easier to stop thinking. Students can get answers instantly, but answers alone don't build problem-solving skills. That's why Koji takes a different approach. Built with learning experts from MIT and Harvard, it helps students work through...
How Healthcare Organizations Can Benefit From an Agentic Artificial Intelligence Workshop
At HIMSS 2026, healthcare IT leaders highlighted agentic artificial intelligence as a top priority, citing NVIDIA’s State of AI report that shows 70% of organizations now use AI and 47% are evaluating agentic AI. Providers face workflow integration hurdles and...
How AI-Driven Workflows Are Changing the Way Companies Think About Data Risk
AI‑driven workflows are reshaping enterprise data‑risk management as firms grapple with exploding data volumes and the need for real‑time threat detection. While 97% of companies invest in big‑data solutions, they typically analyze only about 38% of that data, leaving gaps...
Operationalizing AI at Scale: A Practical Framework for Enterprise-Scale Success
Healthcare AI is moving beyond isolated pilots, with 43% of organizations testing agentic AI, yet only 4% have scaled solutions enterprise‑wide. UiPath proposes a four‑step framework—data ingestion, analysis, entry, resolution—augmented by intelligent document processing, agentic AI, RPA, and a human‑in‑the‑loop....
Singulier and Ommax Merge to Form $90M European AI Consulting Giant
French consultancy Singulier and Germany's Ommax have merged, creating a pan‑European AI consulting platform with over 400 employees and €80‑100 million ($86‑108 million) in revenue. Backed by Eurazeo’s LBO and reinvestments from Founders Future, the new group aims to triple revenue in...
Infineon Teams with NVIDIA MGX AI Factory to Power Next‑Gen AI Server Racks
Infineon Technologies has entered NVIDIA's MGX AI Factory ecosystem, supplying silicon‑carbide and gallium‑nitride power‑delivery components for 800 VDC AI server racks. The move aims to boost compute density while keeping power, cooling and footprint constraints in check.
OpenAI and Anthropic Double Down on Forward‑Deployed Engineering Teams to Embed AI in Production
OpenAI and Anthropic announced accelerated hiring for forward‑deployed engineering (FDE) squads that embed large‑language models directly into customer production environments. The move follows a surge in demand, with job postings for FDE roles up 800% YoY, and reflects a shift...
AI Can Change the World—If We Change Who It’s Built For
In Q1 2026 investors poured $300 billion into 6,000 AI startups, yet less than 10% of that capital targets social and environmental challenges. The article highlights founders like Temie Giwa‑Tubosun of LifeBank, the FLAIR language‑preservation project, and Amini’s SMS‑based farming platform as...
Exasol Launches 2026.1, Branding Its Database as a Sovereign AI ‘Panic Room’
Exasol unveiled version 2026.1, positioning the product as a sovereign AI ‘panic room’ that keeps models and agents inside the database. The release adds native AI functions, an MCP server for controlled model access, and promises up to 1000× faster...
Synergy Quantum Launches SynQ MythGuard as AI-Driven Attacks Outpace Cloud Defenses
Synergy Quantum announced the SynQ MythGuard platform, an AI‑powered solution designed to detect and block Mythos‑class attacks that can locate and exploit vulnerabilities at machine speed. The launch underscores growing alarm among cloud security firms that AI is now outpacing...
No Need for Tickets Says Workday's Jerry Ting, as New Agents Take on IT Provisioning and Travel
Workday unveiled two AI agents—Travel Agent and Sana for IT Service Management—that automate multi‑step workflows across disparate enterprise applications. The Travel Agent can schedule trips, book travel, update calendars, and generate expense reports while enforcing corporate policies. Sana for ITSM...
I'm Starting Something Local (and You're on My List for a Reason)
In this episode, AI advisor Camille Bank introduces the AI Adopters Coffee Club, a local‑focused network where professionals can meet, share AI implementation ideas, and collaborate on projects. She explains how listeners can join a waitlist, either by joining an...
What We Learned From 3 Million Downloads of Kilo Code
Kilo Code announced that its AI‑coding extension has surpassed three million downloads and processed more than 40 trillion tokens across VS Code, JetBrains, CLI, Slack, and Teams. The growth forced the team to evolve from a simple sidebar tool into a full‑stack,...
Funded: Fonoa Raises $110M to Build the Operating System for Autonomous Tax
Dublin‑based fintech Fonoa announced a $110 million Series C round led by Headline to develop an AI‑powered Tax Operating System. The platform consolidates tax‑ID validation, real‑time determination, e‑invoicing and returns into a single, auditable data model. AI agents monitor obligations, populate filings...
Fiserv Turns to Devin AI to Speed Core Banking Upgrades
Finserv has partnered with Cognition to integrate Devin, an autonomous AI software engineer, into its core banking platform. Devin can plan, write, test, and deploy code across complex codebases, dramatically shortening release cycles for bank clients. The initiative aims to...
MeMo's Memory Model Lets Teams Upgrade Their LLM without Retraining It — and Performance Jumps 26%
Researchers introduced MeMo, a modular framework that pairs a small memory model with a frozen executive LLM to ingest new knowledge without retraining the main model. By encoding updates in a dedicated memory model and using model‑merging techniques, MeMo adds...
Hands-On With Gemini Spark: I Gave It Access to My Life and It Friend-Zoned My Boyfriend
Google unveiled Gemini Spark at its I/O conference, positioning it as an always‑on AI agent that taps into a user’s Gmail, Docs, and Calendar to automate tasks. The beta rolls out to subscribers of the $100‑per‑month AI Ultra plan and lives...
Build Your Own GPT-Style Model From Scratch
The post walks readers through building a tiny GPT‑style model from scratch, using open‑source tools like nanoGPT and a modest dataset such as Shakespeare or personal notes. It clarifies three pathways—beginner, practical fine‑tuning, and large‑scale training—emphasizing that the guide focuses...
PNNL Researchers Showcase AI Leadership at AI+ Expo
PNNL scientists highlighted DOE’s Genesis Mission at the AI+ Expo in Washington, D.C., a three‑day event that drew about 20,000 participants, 400 speakers and 175 exhibitors. The lab’s chief AI scientist, Court Corley, presented the Transformational AI Models (ModCon) effort,...
AI Data Centers Now Account for 17% of Australia’s Private Investment
Australia’s AI‑driven data‑center construction surged in Q1 2026, pushing private investment up 6.5% to its highest level in years. Companies poured about A$8.7 billion (≈$6.2 billion) into servers, racks and cooling systems, nearly doubling the previous quarter’s spend. Data‑center‑related projects accounted for 17%...
AI Compute Shifts to Cloud Giants, Redefining Enterprise Strategy
One of the most important and underappreciated realities in AI right now: The frontier labs actually do not control most of the world’s AI compute anymore. The dominant share is increasingly owned by hyperscalers, cloud providers, sovereign AI initiatives, governments, and enterprises...
AI Searches Go Visual, Voice, Camera—Keyword SEO Obsolete
Nearly 1 in 6 searches in AI Mode is now visual, voice, or camera-based, according to @Google. You can’t keyword your way into AI visibility. Instead, you need to prepare for Agentic Visibility. Here’s how: https://t.co/7okFaUCwFB
Will AI Agents Push Enforcement Back Into the Database?
Enterprises testing AI agents that generate and run SQL are confronting a security gap at the data layer. Vendors such as CedarDB argue that traditional role‑based access control and row‑level security must be re‑embedded directly in databases rather than handled...
Treating AI Like A Coworker May Be Making Employees Less Accountable
Boston Consulting Group’s new study reveals that treating AI tools as "employees" erodes human accountability. Nearly one‑third of managers across the U.S., Canada and the EU view AI as a teammate, and over 20% place AI agents on org charts....