Today's AI Pulse
Anthropic shuts down Claude Fable 5 and Mythos 5 after U.S. export‑control warning
Anthropic disabled access to its Claude Fable 5 and Claude Mythos 5 models after the U.S. Commerce Department warned that foreign nationals could exploit the systems. The warning followed alerts from Amazon and other tech firms about a potential jailbreak that could facilitate cyber‑security attacks. Anthropic declined the administration’s request for a voluntary shutdown.
Also developing:
By the numbers: Meta scraps $2B Manus acquisition amid China divestiture order
Deus Ex Machina: Half of US Christians Trust AI's Spiritual Advice
A Barna poll reveals that 48% of practicing U.S. Christians trust AI to aid their spiritual growth, and 34% consider AI advice as trustworthy as a pastor’s counsel. Trust is higher among younger believers, with 39% of Gen Z and 44% of Millennials seeing AI and clergy on equal footing. In stark contrast, only 12% of pastors believe AI can help people grow spiritually. Despite the openness, major concerns persist, as 83% fear AI will misinterpret scripture and 73% worry it could erode faith.
Hyperscience Named a Leader and a Customer Favorite in Document Mining and Analytics Platforms Evaluation
Hyperscience was named both a Leader and a Customer Favorite in Forrester’s Q2 2026 Wave for Document Mining and Analytics Platforms. The company earned perfect 5‑out‑of‑5 scores across 15 Current Offering criteria and three 5‑out‑of‑5 marks in Strategy for Vision, Innovation...
Americans Can’t Spot a Deepfake, and That’s a Business Crisis, Not Just a Consumer Problem
A Veriff‑Kantar 2026 survey of 3,000 adults in the U.S., U.K. and Brazil found Americans score just 0.07 on deep‑fake detection – essentially random guessing. While 63% of U.S. adults recognize the term “deepfake,” half claim confidence in spotting them,...
Cranium AI and Weights & Biases Partner to Make AI Safety and Security a Standard Part of Model Development
Cranium AI announced a strategic integration with Weights & Biases, embedding its security and governance checks directly into the W&B Registry. The partnership lets enterprise AI teams run Cranium safety evaluations on models stored in the registry, with results automatically written...
Google's New AI Search Ads Are Designed to Look Like Helpful Answers
Google is rolling out AI‑driven search ads that appear as conversational answers within its new AI Mode. The ads will be labeled as sponsored but blend with AI‑generated responses, including short product descriptions for shopping queries. Users will still see...
Tamber Unveils $5M‑Backed Sonic Intelligence Suite to Translate Emotion Into Music
Tamber, a Los Angeles‑based creative tech firm, launched its Sonic Intelligence‑powered Creative Suite, a platform that lets musicians convert abstract concepts into sound. Backed by $5 million from Adobe Ventures, M13 and Rackhouse, the suite positions itself as an artist‑first alternative...
ASOS Launches AI Stylist on ChatGPT, Adding Video‑powered Outfit Recommendations
ASOS has integrated its AI Stylist into ChatGPT, offering UK and US shoppers real‑time outfit suggestions with video and livestream previews. The move follows a 52% cut in operating losses to £100.9 million (≈$126 million) and signals a deeper push into conversational...
Neaton Auto Products Manufacturing (NAPM) Improved Quality Control and Reduced Search Time by 70% with CADDi AI Data Platform
Neaton Auto Products Manufacturing (NAPM), a Tier‑1 automotive supplier, deployed CADDi’s AI‑powered data platform to centralize over 140,000 design, procurement and quality documents. The unified repository cut drawing‑search time by 40‑70% and halved QA query duration, saving roughly 1 minute...
Klarna Debuts ChatGPT Shopping App with Live Prices, Tapping AI‑Driven Commerce
Klarna rolled out the Klarna Shopping Search app inside ChatGPT, letting users ask natural‑language queries and receive live product listings, prices and availability from over 100 million items. The move targets AI‑generated traffic that grew nearly 700% during the 2025 holiday...
Camunda Unveils ProcessOS, AI‑First Operating System for Enterprise Workflows
Camunda introduced ProcessOS, an AI‑driven intelligence layer for its agentic orchestration platform, at CamundaCon before 1,200 enterprise leaders. The new system promises to redesign legacy processes around AI, turning recommendations into autonomous actions.
EY and Microsoft Commit $1 B to Scale Enterprise AI for Clients
EY and Microsoft announced a $1 billion joint investment to help clients scale enterprise AI, combining Microsoft’s Azure and Copilot tools with EY’s industry expertise. The partnership aims to embed AI across finance, tax, risk, HR and supply‑chain functions for firms...
Experts Warn AI‑First Coding Tools Threaten Enterprise Accessibility Standards
Mike Paciello, AudioEye's chief accessibility officer, and other experts are urging enterprises to adopt formal accessibility standards for AI‑first development tools. They point to a 95.9% failure rate on top homepages and a surge in costly lawsuits as evidence that...
White House Postpones Signing of AI Executive Order
The White House delayed the signing of a high‑profile AI executive order that would have required national security and civilian agencies to integrate artificial‑intelligence models into federal network defenses and create a voluntary framework for early government review of AI...
Mathematicians Stunned by AI's Biggest Breakthrough in Mathematics Yet
OpenAI’s new artificial‑intelligence model has solved the 80‑year‑old planar unit‑distance problem, a conjecture first posed by Paul Erdős. The breakthrough, announced on 21 May 2026, marks the first AI‑generated proof that experts deem publication‑ready for top journals. Leading mathematicians such as Misha...
Confluent Launches Real‑Time AI Suite to Secure Streaming Data at Scale
Confluent, the data‑streaming pioneer now owned by IBM, rolled out new features in Confluent Intelligence and Confluent Cloud that unify the AI lifecycle, automate privacy controls and enable private connectivity to external models. The upgrades target the security and complexity...
The AI World of Today
Artificial intelligence has moved from novelty to core infrastructure across media, entertainment, and tech. OpenAI’s GPT‑5.3 Codex is being used to self‑improve, while platforms like Moltbook host AI‑only conversations. Major institutions—from Getty Images to the Oscars—have begun banning AI‑generated content,...
Manhattan Associates Unveils AI‑Driven Solution Design Studio to Speed Ecommerce Supply‑Chain Setup
Manhattan Associates introduced Solution Design Studio, an AI‑powered platform that translates natural‑language blueprints into live supply‑chain configurations. The tool promises to shrink implementation cycles from months to minutes, targeting warehouse and transportation managers rather than IT specialists.
AI-Generated Emails Make Business Email Compromise Nearly Undetectable
AI-powered generators are producing business‑email compromise (BEC) messages that mimic real executives and vendors, making detection almost impossible. The FBI reports over $20 billion in internet‑crime losses in 2025, with BEC as the second‑largest method, prompting urgent calls for cost‑free verification...
Smarter Scaling Laws Promise Cheaper, Faster AI Training
Better scaling laws could quietly change the economics of AI. If researchers can predict model performance with far less compute, training frontier systems may become cheaper, faster and less wasteful. The next AI advantage will not only be bigger models, but smarter...
Google DeepMind’s Tulsee Doshi Says AI’s Next Phase Depends on User Trust
Google DeepMind unveiled its Gemini 3.5 models at I/O, powering new personal AI agents, code generators, search tools, and a physics‑accurate video "world model." Product VP Tulsee Doshi explained the tightrope between safety and product quality, highlighting guardrails against harms, sycophancy,...
Google Won't Rest Until Gemini Is Everywhere in Your Home
Google announced Gemini for Home, a full‑stack AI solution that merges Gemini’s large‑model capabilities with Google Home APIs. The program supplies hardware partners with validated reference designs—SoCs, sensors, microphones—to accelerate AI‑enabled cameras and speakers. Google also introduced a Google Home...
Inside AI Tokenomics: How to Profitably Turn Tokens Into Business Value | NVIDIA AI Podcast Ep. 299
In this episode, NVIDIA’s Shruti Kopakar breaks down "tokenomics" – the economics of AI tokens – into four pillars: utility, demand, supply, and monetization. She explains that a token’s value hinges on the intelligence embedded in it (model complexity and context)...
Google Unveils Omni: Personalized Cross‑Device AI Future
Google just revealed Omni, personalized cross-device intelligence, and Spark agents at I/O 2025. I sat down with CEO Sundar Pichai to figure out what comes next: 1:46 Omni: "Nano Banana for video" 4:59 The future of YouTube 7:04 Advice for AI skeptics 9:33 Why your...
Expedia to Launch Agentic AI Tools for B2B Partners
Expedia Group will roll out a Model Context Protocol (MCP) server in the coming months, allowing B2B partners’ AI agents to directly access its travel inventory. The MCP acts as a standardized interface, replacing custom integrations and enabling scalable AI‑driven...
Agentic AI in Financial Services: How Nasdaq eVestment Drives Faster Investment Insights
Nasdaq eVestment has launched an agentic AI solution, the “Next Best Action” agent, built on Snowflake’s AI Data Cloud, to deliver near‑real‑time, AI‑ready institutional data for asset managers. The agent combines a proprietary scoring model with Snowflake Cortex AI to...
Spotify Targets High-Spending Superfans with AI-Generated Music
Spotify is piloting an AI‑generated music service aimed at its highest‑spending subscribers, offering custom tracks created on demand. The feature, built with OpenAI‑backed technology, will be bundled into a premium tier priced around $20‑$25 per month, targeting users who already...
Eagle Vision Launches AI Damage Reporting Tool
Eagle Vision Security unveiled Intelligent Damage Reporting (iDR), an AI‑driven platform that scans trucks and trailers for visible damage in real time. The system combines high‑resolution imaging, video analytics and centralized reporting to automate inspections at logistics yards, terminals and...
US Cyber Command Races to Deploy AI on Top-Secret Networks
U.S. Cyber Command has created a joint task force with the NSA to fast‑track the deployment of commercial AI models—such as those from OpenAI, Google and Anthropic—on the Pentagon’s most classified "high‑side" networks. Led by Gen. Joshua Rudd, the effort...
Turn AI Chat Into a 15‑minute Personalized Daily Brief
Ten months ago, building a truly personalized daily brief from your connected tools wasn’t really practical. Now you can do it in about 15 minutes. This changes what’s possible. A lot of people are still treating AI like a chat...
Brendan Pierson: Illinois’ New AI-in-HR Rules Go Far Beyond Other States’ Restrictions
Illinois enacted an amendment to its Human Rights Act that forces employers to fully disclose any AI tools used in hiring, promotion, or termination decisions and bans discrimination based on protected classes and ZIP codes. The law took effect in...
Indigenous TTS Empowers Language, Keeps Control Local
What happens when a low-resource language builds its own TTS stack and keeps ownership local instead of in Big Tech’s hands? https://spectrum.ieee.org/indigenous-ai-voice-models-maori?share_id=9503722
AI Builds Full Revenue Site in Minutes, Not Weeks
I spun up a full AI revenue agent site in about 2 minutes. Not just a simple landing page. A complete site with a second page, conversion tracking pixels, a smoke test, and a deployed domain. That used to take weeks. Here's...
Māori Text-to-Speech Model Spurns Big Tech’s Values
University of Waikato researchers have built a high‑fidelity text‑to‑speech system for the Waikato‑Maniapoto dialect of te reo Māori using less than eight hours of recorded speech. By leveraging phoneme‑based training and the open‑source Piper architecture, the model achieved a 6.78%...
SAP Taps Mistral AI to Help Customers Migrate Legacy Software
SAP has partnered with Mistral AI to embed large language models into its S/4HANA migration toolkit. The collaboration powers a multilingual retrieval‑augmented generation chatbot that fielded queries from 30,000 Swiss Federal Railways employees, pulling from internal documentation and escalating unanswered...
How AI Is Collapsing the Gap Between Supply Chain Planning and Execution
AI is compressing the traditional divide between periodic supply‑chain planning and reactive execution, creating a continuous decision‑making loop. Forecasts, inventory policies, and transportation plans can now be updated in near real‑time as demand, capacity, and risk signals evolve. The shift...
Google Is Trying to Integrate Too Much AI Too Quickly
At Google I/O, CEO Sundar Pichai unveiled a sweeping AI overhaul that embeds generative models into Search, Gmail, Docs and new agent platforms such as Gemini Spark. The changes replace traditional link lists with interactive, conversational results and add voice‑enabled inbox and...
AI Set to Reshape Secondary Market Pricing and Trading
Artificial intelligence is poised to overhaul pricing and trading in the mortgage secondary market, according to leaders at the Mortgage Bankers Association conference. Executives described AI‑driven scorecards that would price each seller’s loans based on performance metrics, and autonomous agents...
AI Coding Tools Permanently Reshape Software Development
As tools like Claude Code get better, more and more developers are happy to hand off coding tasks to them. The way software gets built has changed for good.
AI Now Turns Ideas Directly Into Published Apps
AnythingAI just made vibe coding way more powerful. You can now ship apps directly to the Google Play Store too. For a lot of builders Apple was the biggest barrier between an idea and a live product. Now AI tools are shrinking that...
AI Just Solved an 80-Year-Old ‘Erdős Problem,’ and Mathematicians Are Amazed
OpenAI announced that its internal large‑language model solved the 80‑year‑old unit‑distance problem, a conjecture posed by Paul Erdős in 1946. The AI generated a high‑dimensional lattice construction that beats Erdős’s best known bound, producing a proof that experts deem clever...
Smarter Workflows, No Code Required: Introducing New AI Steps in Workflow Builder
Slack announced a new Generate AI response step for its no‑code Workflow Builder, letting users embed AI‑driven summarization, translation and drafting directly into automated flows. Users simply add the step, write a plain‑language prompt and attach Slack channels, files or...
Standards, Not Sentences: How to Make Agentic AI Work Across Travel - By Sylvain Roy
Sylvain Roy, CTO of Amadeus, argues that agentic AI will reshape travel only if the industry adopts shared standards rather than isolated solutions. He outlines the Model Context Protocol (MCP) as a first step toward AI interoperability, but notes its...
AI Spend Must Be Tied to Business Value: Tech Leaders at ETCIO Annual Conclave 2026
At the ETCIO Annual Conclave 2026, senior tech leaders stressed that AI spending must be tied to measurable business outcomes rather than sheer volume. They called for stronger FinOps discipline, clear governance, and enterprise‑scale architecture to turn pilots into enterprise‑wide...
ECI Software Solutions Launches Field Service Technician AI Assist Enabling AI Troubleshooting in GlobalEdge
ECI Software Solutions announced AI Assist, an AI‑powered diagnostic feature embedded in the GlobalEdge field service management platform. The tool delivers real‑time, contextual troubleshooting recommendations to technicians via the mobile app, using work order and equipment data. By cutting callbacks...
AI‑Designed PE8 Editors Boost Prime Editing Efficiency
Today in @NatureBiotech we report a new suit of PE8 prime editor proteins. PE8 variants were developed from laboratory-evolved PE6 proteins using AI-guided protein redesign. This approach combines recent advances in computational protein design and directed evolution to increase prime...
Google Pushes AI Agents Despite Uncertain Consumer Demand
Google is pitching an AI agent ecosystem to consumers who may not buy it https://t.co/TRPEXxrkja
Cara Joins Microsoft Teams: Structured Interviews Without Switching Apps
SocialTalent’s AI hiring assistant Cara now lives natively inside Microsoft Teams, letting interviewers pull up interview plans, questions, scoring rubrics and real‑time transcription in a single pane. The integration automatically captures candidate statements, maps them to evaluation criteria, and generates...
No Universal Standards Across LLM Providers Hinder Shared Advice
The shared standards that made one engine’s advice apply to all of them don’t exist between LLM providers. https://t.co/IctQq2LK4k via @DuaneForrester, @sejournal
Network‑Native AI Powers Real‑Time Multilingual Conversations
A strong example of network-native AI in action: @TMobile’s Live Translation beta enables context-aware and real-time conversations across 80+ languages directly through our network. https://t.co/jzkzom6WmS https://t.co/SCW2b3N4iU
Trump Order Pushes Early Government Access to Frontier AI
Scoop: Trump AI executive order seeks early government access to frontier models Already happening... https://t.co/gnyWnGShOv