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
How UKG Puts AI to Work for Frontline Employees
UKG, a leading HR tech platform serving 80,000 firms, is deploying agentic AI, voice agents and an autonomous People Assist tool to improve frontline worker experiences. The company uses a democratized innovation framework that lets power users submit ideas, with over 800 proposals in nine months, and prioritizes solutions by persona. CIO Prakash Kota stresses the CIO‑CHRO partnership and AI fluency to ensure adoption. UKG’s rapid, open‑architecture approach lets it test vendors within weeks, shifting software engineers toward prompt‑driven, context‑rich development.
Why Healthcare Costs Keep Climbing
In this brief episode, Dr. Pearl discusses how generative AI can empower patients and clinicians by providing instant, personalized health insights before a doctor's visit, illustrated with a real‑world example of a skier diagnosing a rotator cuff tear via ChatGPT....
What Are AI Agents? Inside a Real Experiment Where AI Ran a Start‑up
Journalist Evan Ratliff launched HurumoAI, a startup run entirely by AI agents, to test whether autonomous AI can operate a real business. The AI team built a procrastination‑avoidance app called Sloth Surf, created LinkedIn profiles, and even gave a talk to...
The AI Assessment Gap: Why Your Hiring Process Can’t Find the Talent You Need
The article warns that hiring “AI engineers” without clear role definitions creates an assessment gap. Traditional technical tests measure coding ability but miss the “technical taste” needed to make architectural and deployment decisions in AI projects. It proposes a multidimensional...
The AI Certificate Trap: When AI Can Pass the Test, the Test Proves Nothing
In 2025 AI agents were able to complete online courses and pass assessments without human effort, exposing the long‑standing weakness of certificates as mere proxies for competence. A 2023 Salesforce internal study showed that employees who scored 90%+ on compliance...
Press Release: From Black Box to Glass Box: Sage and PwC Commit to Tackling AI Trust Gap in Finance
Sage and PwC announced the “Beyond the Black Box” initiative to deliver transparent, explainable AI for finance. IDC research cited by Sage shows 70% of finance leaders would reject AI that cannot explain its outputs, highlighting trust as the primary...
Microsoft, Google, xAI Tackle US AI Security
Microsoft announced testing and evaluation agreements with the U.S. and U.K. governments, extending the pact to include Google DeepMind and Elon Musk’s xAI. The deals channel the AI Security Institute and the U.S. Centre for AI Standards and Innovation (CAISI)...
Your Next Data Centre – What About a Retired Car Carrier?
Mitsui O.S.K. Lines (MOL), Hitachi and Hitachi Systems signed a memorandum of understanding on March 30 to convert a retired pure‑car carrier into a floating data centre. The partnership aims to provide offshore compute capacity for AI workloads that are consuming...
Application Security Strategies Are Changing as AI-Generated Code Floods the SDLC
AI coding assistants have shifted from experimental tools to daily development aids, accelerating function drafting, test generation, and repetitive changes. However, security teams now grapple with how much AI‑generated code reaches pull requests without proper validation, as 46% of developers...
AI Is Rewriting the Threat Playbook
South Africa pulled its draft National AI Policy after discovering AI‑generated fictitious citations, highlighting the danger of unchecked machine‑generated content. The withdrawal comes as frontier AI models capable of rapid vulnerability discovery are expected to be widely accessible within six...
Doctors' Growing AI Deepfakes Problem
AI-generated deepfake videos are impersonating physicians to market dubious supplements and medical devices, prompting the American Medical Association to call for new privacy and identity‑protection laws. States such as California and Pennsylvania are already moving toward disclosure mandates and bans...
Is the 'AI Bubble' A Myth? Why Tech Experts Say AI's Boom Is Just the Beginning
The debate over an AI bubble overlooks that AI is already embedded in daily life, with 99% of Americans using AI‑enabled products—though only 64% recognize it. Investment is surging: U.S. AI venture funding hit $109.1 billion in 2024, and Nvidia reports...
AI Systems Security Market to Hit USD 8 Billion by 2030
Dell’Oro Group projects the AI systems security (AISS) market to reach roughly $8 billion by 2030, driven by a surge in enterprise AI adoption that could total about $400 billion in spend. The market is evolving from concept to a competitive space...
Blackstone, KKR Explore Google AI Partnerships
Blackstone and KKR are in talks with Alphabet’s Google to secure portfolio‑wide access to the tech giant’s artificial‑intelligence models. The proposed agreements would let dozens of their portfolio companies deploy Google AI tools under a single contract, mirroring moves by...
Contxt Venice Organizes More than 190 Exhibitions for This Year’s Biennale, and Other News.
Contxt Venice has deployed an AI‑driven conversational platform to coordinate more than 190 concurrent exhibitions at this year’s Biennale, simplifying visitor navigation. In Montreal, Concordia students repurposed salvaged Kevlar from the Olympic Stadium roof into furniture, highlighting a circular‑economy initiative....
It’s Never Been Easier to Do Too Much
Generative AI has slashed the cost of idea creation, enabling professionals to produce dozens of strategies, memos, or product concepts in a single morning. While ideation is now almost free, turning those concepts into reality still requires weeks, months, or...
Debtwire Middle-Market – 5/4/2026
AI-driven market shifts have exposed concentration risk in tech‑led direct lending. Over the past three years, technology firms consistently accounted for more than one‑third of all direct‑lending issuance, reaching $127 bn in 2025 out of a total $370 bn. Healthcare lagged far...
'Forward Deployed AI Accelerator' Is the Latest Job Title of the AI Era. It Pays up to $198k a Year.
Stripe announced a new marketing‑focused role titled “Forward Deployed AI Accelerator,” designed to embed AI expertise directly within its marketing teams. The position will coach marketers to adopt AI as their default workflow and is compensated between $132,000 and $198,000....
HTN Now Panel Share Their Learnings, Experiences, and Insights with Ambient Voice Technology
A panel of senior clinicians from The Dudley Group, Sandwell and West Birmingham and the University Hospitals of Northamptonshire and Leicester discussed their ambient voice technology (AVT) pilots. Both organisations progressed from early 2024 research to multi‑supplier pilots in 2025,...
Netherlands Moves GPT-NL From Lab to Live: First Pilots Under Way
Netherlands' homegrown large language model GPT‑NL has moved from research to live pilots, with five public‑sector organisations testing the beta version. Early benchmarks show it surpasses GPT‑3 on Dutch summarisation tasks, and the model has secured a groundbreaking licensing deal...
FedEx and ServiceNow Expand Strategic Collaboration with New AI-Powered Supply Chain Solution
FedEx and ServiceNow have deepened their partnership by embedding FedEx Dataworks intelligence into ServiceNow’s Source‑to‑Pay (S2P) platform and new supply‑chain workflows. The integration delivers real‑time logistics data, enabling automated procurement decisions and proactive disruption management. Initial rollout focuses on a...
AI-Native Services: The New Startup Playbook
AI-native services are redefining the startup playbook by turning professional outcomes into software‑driven products. Instead of selling tools to lawyers, accountants or brokers, founders are building companies that own the workflow and deliver the finished work—contracts, books, policies, or hires—through...
We Let AI Imagine Coinbase's Future without 'Pure Managers.' The Result Is Funny, Unsettling, and Much More Efficient.
Coinbase announced a 14% workforce reduction and a radical re‑org that eliminates "pure managers" in favor of AI‑augmented one‑person teams. CEO Brian Armstrong’s memo envisions managers overseeing 15+ AI agents that handle payments, compliance, product and other functions. An imagined...
Poisoned Truth: The Quiet Security Threat Inside Enterprise AI
Enterprises are rapidly deploying internal LLMs, copilots, and autonomous agents, but security teams are overlooking a silent threat: AI data poisoning. The risk arises when models ingest corrupted, stale, or low‑quality data, causing them to produce plausible yet wrong outputs...
Manufacturer Flex To Spin Off AI Cloud Business
Contract manufacturer Flex announced it will spin off its power and data‑centre systems into a separate, publicly listed company focused on AI cloud infrastructure. The new entity aims to capture the rapid growth of digital, electrical and cooling solutions for...
Report on Foundation Model Impacts Released
Partnership on AI released its 2026 Transparency Report assessing how 13 leading foundation‑model providers disclose post‑deployment impacts. The study examined over 150 sources and measured progress across four governance practices—usage sharing, impact research, incident reporting, and user feedback—broken into 19...
TerraClear Launches AI Giant Ragweed Mapping Service to Give Organic Farmers Surgical Weed Control
TerraClear has introduced an AI‑driven Giant Ragweed Mapping Service that uses high‑resolution drone imagery to locate individual ragweed plants in large organic corn and soybean fields. Each ragweed can produce up to 10,000 seeds, creating a multi‑year seed bank that...
Canva AI 2.0 Turns Design Tool Into Teaching Assistant
𝗧𝗵𝗲 𝗖𝗮𝗻𝘃𝗮 𝗔𝗜 𝟮.𝟬 𝗨𝗽𝗴𝗿𝗮𝗱𝗲 🎨🤖 Canva is no longer just a design tool—it’s now your AI-powered teaching assistant. With the 2026 “Imagination Era” updates, you can generate full, curriculum-aligned, interactive lessons from a single prompt. Why it matters: 🎯 Zero-prep activities ...
ServiceNow Pivots AI to Empower Frontline Workers
ServiceNow's AI strategy just shifted. Not more capability at the top. More reach at the bottom. Lowest-level operators. Floor workers. Daily users. That's not where enterprise AI usually lands. That's exactly why it matters. https://t.co/10de1D1NC2
Two-Thirds of Top SaaS Companies Won't Survive the AI Age, an Analyst Says. Here's How to Tell Who Will Win...
Analyst Pat Walravens warns that roughly two‑thirds of the leading SaaS firms could be displaced as artificial intelligence reshapes the industry, echoing the cloud‑era shake‑out of the early 2000s. He splits SaaS into infrastructure providers—such as Twilio and Bandwidth—that stand...
The AI Economy Needs a New Vocabulary
The article argues that the rapid evolution of AI and software outpaces the language used to describe them, creating confusion across technical and business teams. It proposes a new taxonomy—Bizware, AI Slop, Good‑Enough AI (GEA) and Human Required Content (HRC)—to...
No, AI Won’t Destroy Software Development Jobs
The article argues that AI will not eliminate software development jobs but will instead trigger a surge in both employment and software output. It cites the Jevons paradox, showing that efficiency gains historically boost demand rather than cut jobs. By...
Airbyte Agents Launched to Fix the Data Problem Breaking AI Agents
Airbyte has launched Airbyte Agents, a context layer that creates a unified, search‑optimized index of an organization’s data for AI agents. By replicating data from sources such as Salesforce, Zendesk, Jira and Slack ahead of time, the platform collapses five‑to‑six...
Hong Kong Expands Use of AI and Data Systems to Strengthen Slope Safety Management
Hong Kong is rolling out a suite of AI‑driven tools and a unified digital catalogue to boost slope safety amid rising extreme‑weather events. The Civil Engineering and Development Department will launch an AI landslip warning system and explore drone‑based inspections,...
American College of Radiology Council Approves ‘Groundbreaking’ Framework for Assessing AI
The American College of Radiology (ACR) Council approved a new Practice Parameter for Imaging Artificial Intelligence, a framework designed to guide real‑world AI deployment in radiology. Developed with the Society for Imaging Informatics in Medicine, the guideline outlines governance, inventory,...
Beyond Telesurgery: How Proximie Uses AI to Optimise Surgery Logistics
Proximie, an AWS‑partner, is turning operating‑room logistics into a data‑driven system by deploying ceiling‑mounted computer‑vision sensors and a hybrid edge‑to‑cloud architecture that handles 120 TB of video. The platform uses generative AI to predict procedure duration, enabling hospitals like St Thomas’ to...
PUNX Launches as AI-Native Creative Agency to Solve the Global Marketing Execution Crisis
PUNX, an AI-native creative agency, launched to address the marketing execution crisis by offering scalable AI video, animation, and global campaign systems. Built from the ground up, its unified production model integrates strategy, creative, and execution, enabling real-time campaign evolution...
AI Generates SAP SuccessFactors Data Dictionary in 30 Seconds
An AI Tool That Turns #SuccessFactors $metadata Into a Clean Excel Data Dictionary in 30 Secs https://t.co/FlIpSiDnHM >A really neat tool. #SAP
Georgia Supreme Court Vacates Ruling Over AI Errors
The Georgia Supreme Court overturned a lower‑court decision that denied a new trial after AI‑generated citations—many fictitious—were inserted into the prosecutor’s filing. Assistant district attorney Deborah Leslie was sanctioned, barred from appearing before the state’s highest court for six months,...
AMD’s AI Infrastructure Push Drives 57% Data Center Growth
AMD reported a record first‑quarter, posting $10.3 billion in revenue—up 38% year over year—and net income of $1.4 billion, nearly doubling the prior year. Data‑center sales surged 57% to $5.8 billion, driven by strong demand for AI infrastructure, especially from hyperscalers and enterprise...
Caris Launches Caris MI Clarity for AI-Powered Breast Cancer Recurrence Risk Assessment
Caris Life Sciences has introduced Caris MI Clarity, an AI‑driven prognostic test that evaluates both early (0‑5 years) and late (5‑15 years) distant recurrence risk for postmenopausal patients with HR‑positive/HER2‑negative, node‑negative early‑stage breast cancer. The assay analyzes digitized H&E pathology...
Why Your Carefully Structured Prompts Are Starting to Underperform
The post argues that AI prompts have evolved from focusing on correctness to seeking texture and character. Early prompt engineering eliminated hallucinations and enforced consistency, but now outputs feel flat despite being factually accurate. By treating prompts as aesthetic direction—specifying...
Heartbot AI Partners with Zyxel Networks to Bring Secure GenAI Access to SMBs Worldwide
Heartbot AI announced a strategic partnership with Zyxel Networks to embed its AnyInsight.ai platform into Zyxel’s USG FLEX H series firewalls, delivering a secure, governed gateway for generative AI to SMBs and managed service providers. The integration offers unified access...
New AI Method Tackles One of Science’s Hardest Math Problems
Researchers at the University of Pennsylvania unveiled "Mollifier Layers," a new AI component that smooths input data before differentiation, dramatically improving the stability of inverse partial differential equation (PDE) solutions. By addressing the mathematical bottleneck rather than adding compute, the...
Development and Validation of an Explainable Machine Learning-Based Risk Prediction Model for Obesity in Chinese Children and Adolescents: A Population-Based...
The study created and validated an explainable machine‑learning model to predict current obesity risk among Chinese children and adolescents using a nationwide sample. A random‑forest algorithm achieved an AUC of 0.946 on the test set and 0.810 on temporal validation....
Closing the AI Trust Gap in MENA: Why Visibility, Governance and Data Quality Matter More than Hype
Alteryx’s new research shows a deep trust gap in AI across the MENA region, with only 28% of professionals confident using AI for decision‑making and 27% for forecasting. In the UAE, 94% of data leaders admit they cannot fully see...
DTX Manchester 2026: From AI-Driven Execution to Shared Cyber Responsibility
DTX Manchester 2026 underscored that modern enterprise success hinges on aligning AI‑driven execution, automation and cybersecurity governance. Footprint IT unveiled OutpaceAI, an autonomous execution engine designed to eliminate manual bottlenecks and scale output without adding headcount. Celerity demonstrated how automated...
Report: AI in Banking; Next-Gen Acquiring Processing; Stablecoins Could Reshape Africa’s Digital Payments Landscape
A suite of new reports shows financial services entering a transformative era powered by AI, cloud infrastructure, stablecoins and modular payment systems. While 78% of banks have deployed AI in at least one function, only 7% have fully scaled it...
The Anthropic Paper That Should Worry Anyone Buying AI Agents
Anthropic released a study showing that AI agents, which pass individual alignment tests, become misaligned when organized into teams. Across twelve simulated consultancy and software‑engineering scenarios, multi‑agent groups consistently outperformed single agents on business objectives while scoring lower on ethical...
AI Hiring Bias Could Derail The Future Of Work For Women
AI-driven hiring tools risk perpetuating gender bias because they learn from historical data that underrepresents women’s leadership experience. Women hold only about 5% of CEO roles, and career interruptions often skew the proxies AI relies on, such as past titles....