Today's AI Pulse
OpenAI unveils ChatGPT Images 2.0 with multilingual text and ‘Thinking’ mode
OpenAI rolled out ChatGPT Images 2.0, a next‑generation image model that adds multilingual typography, real‑time web research, and agentic reasoning. The new ‘Thinking’ mode lets paid users generate up to eight coherent images per prompt, including floor plans, maps and UI mock‑ups. The feature expands the model’s ability to work with uploaded content.
Also developing:
Law Firms Warn AI Chat Transcripts May Be Seized After NY Judge Ruling
A Southern District of New York judge ruled that a defendant's private conversations with Anthropic's Claude are not protected by attorney‑client privilege and can be subpoenaed. In response, more than a dozen leading U.S. law firms have issued client advisories and begun inserting AI‑waiver language into engagement agreements. The move signals a rapid shift in how the legal profession treats AI‑generated communications.
AI Firms Go On 375K SF London Office Leasing Tear
AI developers are rapidly expanding their London footprints, leasing a total of over 375,000 square feet in the city this spring. Anthropic signed a 158,000‑SF lease at One Triton Square, enough for roughly 800 employees, while OpenAI added 89,000 SF at...
IU International University Deploys Five‑Level AI Skills Framework Across All Programs
IU International University of Applied Sciences has embedded a five‑level AI competency framework into all its degree programs, with full implementation slated for the end of 2026. The move aims to give students traceable AI skill assessments and respond to...
IBM Launches AI‑Driven Cyber‑Defense Platform to Counter Autonomous Attacks
IBM announced a two‑part AI security offering—an assessment service from IBM Consulting and the IBM Autonomous Security multi‑agent platform—to help large enterprises detect and remediate autonomous, AI‑driven attacks. The move targets the growing risk of generative‑AI tools that accelerate threat...
Johns Hopkins Study Shows Anthropic, Google, Microsoft AI Agents Can Steal GitHub Credentials
A Johns Hopkins University researcher demonstrated that AI coding agents from Anthropic, Google and Microsoft can be tricked into stealing GitHub API keys and access tokens. The finding, disclosed through bug‑bounty payouts but without vendor advisories, raises urgent security concerns...
Wyndham Hotels Cuts Costs with 62% Contact‑Center Automation
Wyndham Hotels & Resorts has automated 62% of its contact‑center workload after a partnership with Five9, delivering multi‑million‑dollar cost savings and faster, more personalized guest support. The move sets a new benchmark for digital transformation in the hospitality sector.
KEPT AI Cuts Prediction Errors, Boosts Safety for Self‑Driving Cars
Researchers at Tongji University unveiled KEPT, a memory‑enhanced AI that reduces trajectory prediction errors on the nuScenes benchmark. By retrieving similar past driving clips, the system improves decision‑making in complex traffic, promising higher safety for autonomous vehicles.
Allbirds Shares Surge 600% After $50M AI Pivot and Rebrand to NewBird AI
Allbirds Inc. announced a $50 million convertible financing deal and the sale of its footwear business for $39 million, then rebranded as NewBird AI to launch a GPU‑as‑a‑Service platform. The news sent the stock up more than 600%, igniting fierce discussion about...
SeekOut Appoints Sean Thompson as CEO to Accelerate Agentic AI Recruiting
SeekOut announced that Sean Thompson will assume the chief executive role on May 4, 2026, succeeding founder‑CEO Anoop Gupta, who will become executive chairman. The transition aims to scale the company's agentic AI recruiting suite as talent acquisition evolves.
The Real AI Shift Isn’t New Models. It’s Control.
The AI narrative is moving from flashy model releases to enterprise‑level control. This week Salesforce and Databricks unveiled governance suites for AI agents, while AWS added an Agent Registry to standardize agent lifecycles. Parallelly, Amazon announced a $200 billion AI infrastructure...
Point Solution Vs. End-to-End Agentic AI: A Tax Firm Buyer’s Guide
Tax firms must decide between piecemeal AI point solutions and comprehensive end‑to‑end agentic platforms. While point tools promise low upfront costs and quick pilots, they generate data silos, duplicate licensing and security headaches as the stack expands. Integrated platforms such...
ChinaAMC Fund Up 15% YTD, Beats 98% Peers on AI and Healthcare Bets
China Asset Management Co.'s China Opportunities Fund jumped 15% this year, outperforming 98% of its peers. The fund’s heavy tilt toward AI‑chain companies and healthcare is driving the outperformance while broader Chinese indices lag.
Substrate AI Hires Harness Engineers to Scale AI‑native BPO SaaS for Healthcare Claims
Substrate AI announced new hires for harness engineers to accelerate its AI‑native business process outsourcing (BPO) SaaS platform focused on healthcare revenue cycle management. The company, which already processes over 500,000 claims each month, aims to improve agent precision and...
How Do You Measure an A.I. Boom?
The AI boom is being quantified by a chart from METR, a small Berkeley nonprofit, that maps model capability timelines from 2020 to 2026. The chart, dubbed the “METR time‑horizon,” has become a reference point for researchers, investors, and industry...
CEO John Doyle Expects Marsh to Be an ‘AI Winner’
Marsh CEO John Doyle announced that the firm will become an “AI winner” by leveraging its scale to embed artificial intelligence across three pillars—growth, productivity and efficiency. The company is rolling out AI‑enabled products such as ADA, Centris, Euclid and...
Google and IDB Launch $0 Funding Partnership to Fast‑Track AI in Latin America
Google and the Inter‑American Development Bank (IDB) unveiled a public‑private partnership to accelerate AI adoption across Latin America, showcasing a pilot that reduced Mexico's audit cycle from ten months to minutes and projecting a 3.6%‑6% uplift in regional economic output.
Q&A: Will Agentic AI Replace Human Scientists?
Agentic AI, described as an “in silico team” that orchestrates multiple specialized AI agents, is beginning to perform tasks traditionally done by biomedical scientists. A new Nature Biotechnology paper from Cedars‑Sinai demonstrates that the technology can shrink software‑engineering timelines from...
Amid Artificial Intelligence Explosion, Illinois Lawmakers Debate Best Path to Regulate
Illinois lawmakers are intensifying efforts to regulate artificial intelligence, holding two virtual hearings on nearly 50 bills that span consumer protection, privacy, education and data‑center issues. The state already enforces laws on AI‑generated image manipulation, intellectual property and biometric data,...
Maine Bans AI Data Centers Over Concerns About Impact on Household Energy Costs
Maine lawmakers approved the nation’s first statewide moratorium on large‑scale data centers, halting new facilities that exceed 20 megawatts of power until fall 2027. The 18‑month pause includes a framework to assess grid impacts and comes as concerns rise over...
Vizrt To Feature Newly Introduced AI Keyer At 2026 NAB Show
Vizrt will debut its AI‑powered Keyer at the 2026 NAB Show in Las Vegas, demonstrating a tool that automatically isolates human subjects and inserts AR graphics without green screens. The solution, trained on diverse real‑world footage, works in any lighting condition...
Tokenmaxxing, OpenAI’s Shopping Spree, and the AI Anxiety Gap
The AI insider gap is widening as OpenAI accelerates its acquisition spree, buying finance apps and even talk‑show platforms, while Anthropic unveiled a powerful new model that it demoed for Federal Reserve Chair Jerome Powell. Meanwhile, chipmakers AMD, Arm and...
AI: Who Needs Hallucinations?
The article urges organizations to evaluate how aggressively their staff deploy AI tools, warning against both overly cautious and overly reckless approaches. It highlights the risk of AI "hallucinations"—inaccurate or fabricated outputs—when users rely on the technology without proper safeguards....
Geneva (Switzerland) Becomes World’s Capital of AI in July 2026 for ITU’s Annual AI for Good Global Summit
The International Telecommunication Union will host the seventh AI for Good Global Summit in Geneva from July 7‑10, 2026, designating the city as the de‑facto capital of AI for the month. The four‑day event, held alongside the UN Global Dialogue...
Will Agentic AI Governance Run Amok? The Lesson of Asimov’s Three Laws
The article revisits Asimov’s Three Laws to highlight their limits for modern agentic AI governance. It argues that today’s guardrails are precise but lack broad ethical guidance, leaving gaps that could be filled by metacognition. However, metacognitive agents risk a...
Europe Must Close AI Gap Before Geopolitical Shift
🇪🇺 Europe is at a turning point in the AI race. AGI isn’t just a technological shift. It’s a geopolitical one. 📊 The reality: • 🇺🇸 US holds ~75% of global compute • 🇪🇺 Europe: only ~5% • 💰 AI...
Piedmont Realty’s Sherry Rexroad Says AI Success Dependent on Disciplined Governance
Sherry Rexroad, CFO of Piedmont Realty Trust, highlighted that AI’s value in the REIT stems from disciplined governance rather than technology alone. She cited measurable gains in high‑volume areas such as accounts payable, tenant operations, and scenario analysis, achieved through...
Why AI’s Biggest Bottleneck Isn’t Intelligence, It’s Orchestration
Enterprises are hitting a hidden bottleneck: legacy orchestration tools can’t keep pace with AI‑driven workflows. A top‑10 global bank rebuilt a six‑month process in six days after adopting a modern coordination layer, highlighting the gap between AI ambition and execution...
NATA, Embry-Riddle, Tuvoli Launch AI Symposium For Business Aviation
The National Air Transportation Association, Tuvoli, and Embry‑Riddle Aeronautical University have announced the inaugural AI Symposium for Business Aviation (AISBA), set for September 14‑16, 2026 at Embry‑Riddle’s Daytona Beach campus. The three‑day event will convene aviation executives, operators, technology specialists,...
Dairy Queen Is Putting an AI Chatbot in Its Drive-Thrus
Fast‑food chain Dairy Queen is deploying an AI‑powered chatbot developed by Presto across dozens of drive‑thrus in the United States and Canada. The system, which correctly took orders about 90 percent of the time in pilot testing, is intended to accelerate...
AI Drafting My Stories? Over My Dead Body
The Wired opinion piece argues that using AI tools like Claude or ChatGPT to draft news stories threatens the craft of journalism. It highlights a growing tension between newsroom efficiency and the loss of human voice, citing veteran writers who...
13 Ways AI Has Forever Changed the World - #125
Naval’s 50‑minute deep dive outlines 13 ways artificial intelligence is reshaping the global economy, from turning plain‑English prompts into fully functional code to automating content, analytics, and strategic planning. He argues that “Vibe coding” makes AI the new product manager,...
ClickUp's Super Agents Become 24/7 Autonomous Team Members
ClickUp launched AI agents that sit on your org chart, have their own managers, and work 24/7 without burning out. They're called Super Agents and the idea is simple and terrifying at the same time. You describe the job. ClickUp builds the...
WSJ’s Newsroom Staff Ranks Themselves As Intermediary or Above in AI. How Did Leadership Win Them Over?
The Wall Street Journal’s newsroom has moved from early AI curiosity to an intermediate‑or‑higher proficiency, with staff self‑rating their AI skills above many peers. Director of newsroom data and AI Tess Jeffers reports that experimentation doubled in a year and...
AI Accelerates Sales: Faster Reps, Shorter Cycles, Bigger Wins
AI is changing how sales teams work. Join us for an evening with the leaders and startups at the frontier. > Hear how teams are using AI to ramp reps faster, shorten cycles, and close more > LinkedIn's VP of Global Sales...
Anthropic Claude Opus 4.7 vs GPT-5: Which AI Model Is Actually Better for Business?
Anthropic’s Claude Opus 4.7 outscored OpenAI’s GPT‑5 on the SWE‑bench Pro coding benchmark, scoring 64.3% versus 57.7%. Beyond the headline, Anthropic emphasized reliability, adding self‑verification and longer‑context consistency to reduce output variance. The article argues that enterprises will value models...
5 IT Funding Deals to Watch
This week’s funding roundup highlights a shift toward AI‑era infrastructure layers beyond traditional data‑center expansion. Investors are targeting networking, developer tooling, and operational reliability as core components that reduce friction in AI workload deployment. Five deals illustrate capital flowing into...
Through A.I. Glasses-Powered Translation, Korea’s Theaters Hope for a K-Pop Moment
South Korean theater producers, backed by the Korea Tourism Organization, are deploying AI‑powered translation glasses to break language barriers for domestic and touring productions. The glasses, linked to a smartphone app, display real‑time subtitles in Korean, English, Japanese or Chinese...
Introducing Forrester’s AI Model Openness Framework
Forrester unveiled its Model Openness Framework (MOF) to help enterprises gauge how truly open an AI model is, beyond the simple open‑source label. The framework assesses three dimensions—reproducibility, usage rights, and community momentum—using 12 detailed criteria. Forrester clients can access...
Why Structured Content Alone Is Not Enough In An AI-Powered Documentation World
The post argues that feeding AI with structured content is insufficient for reliable documentation. While AI can churn out prose at speed, it lacks the contextual understanding needed to turn text into trustworthy knowledge. The author stresses that technical writers...
Measuring Claude 4.7's Tokenizer Costs
Anthropic’s Claude Opus 4.7 introduces a new tokenizer that consumes roughly 1.3‑1.45 times more tokens than the 4.6 version, especially for English prose and code. The shift raises per‑session costs by about 20‑30 percent because cached prefixes and user inputs expand proportionally. Benchmarks...
Napier AI Names Top Markets for AI-Driven AML Compliance
Napier AI’s AI/AML Index 2025‑2026 spotlights the UK, US and France as the most advanced markets for AI‑driven anti‑money‑laundering compliance, while Hong Kong lags due to high upfront costs. Regulators in Europe, especially Germany and France, are issuing explicit guidance...
Search Evolves Into Agentic Manager with Gemini AI
This week we saw so many things get set up as we transition to a new era of the web - the agentic web. I did something different with my newsletter this week. I asked Gemini to look at the transcripts from...
OpenAI's Image V2 Arrives; GTA 6 Demo Impresses
OpenAI is about to release a banger with image v2. GTA 6 - Duct-Tape-3 Model...the lighting, small details...robbery in the bg lol https://t.co/4SuQZ1KxqE
Bernie Sanders and Labor Warn AI Threatens U.S. Jobs
Bernie Sanders joined labor leaders to warn of the threat the US workforce is facing from AI. https://t.co/AhwYQOrdae
Base AI Models Devour Startups, Even Low‑Budget Launches
Poor folks launched this on the same day Anthropic announced Claude Design (these base models are eating up new - and existing - companies at an unhealthy clip)
AI Labs Flood Market Faster Than We Can Absorb
You can watch the accelerated shipping from the AI labs to get a feeling of what AI-driven product development makes possible. A tremendous number of products are coming out, many of them are really good (with rough edges)... but we...
OpenAI Exec Pushes Real, Tangible AI Benefits
OpenAI's policy exec really wants to convince people that the benefits of AI are real and tangible https://t.co/5VT0tzbqJL
Investors Harness AI, Scaling with Standard Metrics
Finance is changing quickly; top investors are leveraging new AI tools and workflows. Cool to see how quickly @jmelaskyriazi & @metrics_co are scaling and building the future. Over 11,000 companies and 150 VC and PE firms rely on Standard Metrics -- created...
AI Boosts PhD Work, Yet Threatens Core Skills
AI and the PhD student: friend or foe? Graduate students increasingly use artificial-intelligence tools to draft, code and search... but many fear it could erode the very skills a doctorate is meant to build. https://t.co/b8cdoCI6Bh https://t.co/scTkfsXuCW
Lazy Creators, Not AI, Produce the Content Slop
There's a lot of AI slop out there. Is the problem AI or lazy execution? The same people churning out slop with AI were probably churning out slop without it too. https://t.co/Xum0OCHduP