AI NEWS YOU MISSED ❗#71

Top 5 stories of the week

1️⃣ AGI Tensions Strain Microsoft–OpenAI Alliance

Microsoft and OpenAI are locked in high-stakes negotiations over future revenue sharing, AGI definitions, and IPO plans — battling over who controls next-gen AI and whether their once-symbiotic relationship can survive competing ambitions and growing mistrust. The Information

2️⃣ OpenAI Designs Rival to Office Workspace 

Yahoo Finance reports that OpenAI is developing its own office workspace software, positioning itself as a direct competitor to established enterprise ecosystems from Google and Microsoft. Yahoo Finance

3️⃣ OpenAI Fights Meta Talent Raid

As Meta lures away top OpenAI researchers with massive offers, OpenAI leadership vows to recalibrate compensation and defend its talent, calling Meta’s tactics a “break-in” and urging staff to stay focused on the long-term AGI mission. Wired

4️⃣ What’s ex-OpenAI CTO Murati doing?

Ex-OpenAI CTO Mira Murati’s startup, Thinking Machines Lab, has raised $2B to build AI models customized to business KPIs using reinforcement learning and model merging, plans to leverage open-source models to accelerate development, and is drawing interest from cloud giants and potential acquirers like Meta. The Information

5️⃣ Copyright Lawsuits Favor AI Firms

In a landmark decision for the AI industry, a federal judge has largely sided with Anthropic (and similarly with Meta) in lawsuits brought by authors, ruling that training AI on rightfully-aquired copyrighted works can be considered fair use, though the legal battles are expected to continue. TechCrunch

INSEAD AI 50 — Nominations Open

More INSEADers Are Building in AI Than You Think. Time to Spotlight Them.

When Fabio, Amit, and I kicked off the INSEAD AI community, one thing became clear fast: a lot more INSEADers are doing serious work in AI than most people realize. Let’s change that.

🚀 INSEAD AI 50 is launching — a curated list of the most impactful AI builders, leaders, and thinkers from across the INSEAD world.

👥 Alumni, students, faculty, staff, researchers ✅ Based on real impact, not votes 🔤 Alphabetical, curated independently 🙋‍♂️ Self-nominations welcome 📝 Nominate as many as you like

Nominate here → 

Upcoming live events — INSEAD-led

Inside Adya: INSEAD Alum Dives Into the Future of AI Agents (click to RSVP)

Join INSEAD alum Shayak Mazumder, CEO & CTO of Adya, for an exclusive Zoom session exploring how AI agents and third-party tools like Make are reshaping business tech, with live insights on Adya’s full-stack AI vision and a dynamic Q&A.

How Policy is Shaping the Future of AI (24 July) (Join the private INSEAD Tech Alumni group)

Join the next INSEAD TechAlumni webinar featuring AI governance expert Nora von Ingersleben-Seip for a timely briefing on how global policies are shaping the future of AI — and what tech leaders need to know to stay ahead in 2025.

INSEAD AI Venture Lab Founder Sprint

INSEAD’s new AI Venture Lab launches the AI Founder Sprint — an 8-week program helping early-stage founders build smarter, faster startups using AI, backed by cutting-edge research and a global network.

📝 Apply to join as a founder — and bring your idea to life with AI

👩 Be a Mentor — guide and inspire the next wave of entrepreneurs

🧭 Be a Venture Guide — help build a global community of founders

🤝 Partner with INSEAD AI Venture Lab— and help scale this global platform

INSEADer News

Phanish Puranam Wins 2025 Oskar Morgenstern Medal

INSEAD’s Prof. Phanish Puranam is being honored for his pioneering research on AI’s impact on strategy and organizational design, with a focus on human–AI collaboration, managerial hierarchy, and human-centric adoption.

INSEADer Tim Gordon with Prof. Theos Evgeniou on AI at INSEAD Reunion INSEADer Tim Gordon spoke alongside Professor Theodoros Evgeniou at the alumni reunion about the opportunities and risks of the Age of Intelligence, sharing insights on AI’s impact on business, individuals, and geopolitics — an especially meaningful moment as he addressed his own promotion and others.

Age of Intelligence EP02 out — with Morgan Stanley

Episode 2 of the Age of Intelligence podcast features Jeff McMillan, Head of AI at Morgan Stanley, sharing front-line insights on scaling AI in traditional industries and navigating the global AI race — essential listening for executives and decision-makers.

APPLE PODCASTS SPOTIFY DEEZER:

More AI stories of the week

Suggested by INSEAD-ers. From the 3,000+ person global INSEAD AI private business leader community, regularly sharing AI-related articles, vids, etc, of interest. (alum-led)

Cheat Sheet — Best AI Apps by Use — June 2025

• 22 New Jobs AI Will Create A NYT Magazine piece by Robert Capps explores how the rise of AI is spawning entirely new roles across Trust, Integration, and Taste — ranging from AI Ethicists and Translators to Personality Directors and World Designers — reshaping the future of work.

• OpenAI Turns to Google’s AI Chips to Power Its Products In a significant strategic shift, Reuters reports that OpenAI is incorporating Google’s tensor processing units (TPUs) to power its models, highlighting the immense compute demands and evolving alliances in the AI infrastructure race.

• Meta Considers Moving Away from Llama and Open Source According to reports, Meta is contemplating a major strategy change by potentially moving away from its open-source Llama models in favor of a new, closed system, a move that could significantly alter the open-source AI landscape.

• Meta Hires Three OpenAI Researchers Amid Talent War Reuters confirms Meta has hired three researchers from OpenAI, underscoring the fierce and high-stakes “talent war” for top AI researchers among big tech firms.

• AI Agents and “Sabotage-as-a-Service”: Anthropic on Agentic Misalignment Anthropic’s latest research paper explores how autonomous AI agents can learn deceptive strategies, including blackmail and strategic deception, introducing the critical business risk of “agentic misalignment.”

• Sam Altman: The Future of OpenAI, ChatGPT’s Origins, and Building AI Hardware (YouTube) In a candid fireside chat, Sam Altman reflects on OpenAI’s journey, the origins of ChatGPT, the road to AGI, and why building great AI — and hardware to run it — requires vision, resilience, and the right people.

• Elon Musk: Digital Superintelligence, Multiplanetary Life, How to Be Useful (YOUTUBE) In this wide-ranging fireside chat, Elon Musk shares lessons from his startup journey, the principles behind SpaceX and Tesla, and his vision for AI, digital superintelligence, and a multiplanetary future — all rooted in doing things that are truly useful.

• New KPI Map Helps Fix AI ROI Gaps IMD has released a multilingual, interactive toolkit mapping digital and AI transformation KPIs — organized by business goals and backed by real-world cases — to help organizations overcome poor measurement and finally see ROI from their AI investments.

Product, Technology & Market Developments

• Claude Desktop Extensions Make AI Setup One-Click Simple Anthropic has launched Desktop Extensions — a new format that lets users install powerful local MCP servers for Claude Desktop with a single click, eliminating the need for developer tools or manual configs and opening the door to wider adoption, enterprise deployment, and creative new use cases.

• Perplexity Labs Turns Ideas into Interactive Projects and Apps Perplexity Labs, now available for Pro users, allows anyone to transform ideas into fully realized projects — ranging from research reports to dashboards and interactive mini-apps — using AI-driven workflows that combine search, code execution, and visual creation, enabling results in minutes that would previously take days.

• Ex-Harvard Professor Claims His AI Can Predict the Next War An NDTV report details the work of a former Harvard professor whose AI model analyzes vast datasets to predict and potentially help prevent future military conflicts.

• The Future of Job Posts is AI-Centric Quora CEO Adam D’Angelo shared a post illustrating how future job descriptions will likely be framed around an individual’s ability to effectively leverage and manage specific AI models and tools.

• Germany Moves to Block DeepSeek Over Unlawful China Data Transfers In a major escalation, Germany has ruled DeepSeek’s data transfers to China unlawful and urged Google and Apple to block the app — marking the first EU country to act against the Chinese AI challenger and signaling how data protection laws are shaping the AI race in Europe.

• Google Releases Multimodal Gemma-3N for On-Device AI (Shared by Benjamin Latli) Google has launched Gemma-3N, a powerful and efficient new multimodal model designed for on-device applications, positioning it as a key component for the next generation of smart devices.

• Interactive Shopping with Real-Time Generated Content A demonstration of how AI can generate interactive, real-time content is providing a glimpse into the future of the e-commerce shopping experience.

• The Monster Inside ChatGPT: AI as a “Shoggoth” A Wall Street Journal opinion piece discusses the concept of LLMs as “Shoggoths” — powerful, shapeless entities whose inner workings remain incomprehensible even to their creators, posing unique safety and alignment challenges.

• Rethinking SaaS Playbooks for the AI Era Andreessen Horowitz shares 5 key principles reshaping enterprise AI startups — from faster ARR growth and collapsing build costs to the rising importance of speed, workflow integration, and moats in a market where flashy demos aren’t enough.

• How to Automate Sales with $0 Tools (YOUTUBE) Grady Teske breaks down how to fully automate your B2B sales process using free AI tools like N8N and ChatGPT — turning traditional field sales tactics into scalable digital workflows that qualify leads, personalize outreach, and drive human connection at scale.

• Denmark Moves to Ban AI Deepfake Abuse Denmark is set to become the first European country to amend copyright law to protect individuals’ voices, faces, and bodies from AI-generated deepfakes, granting people the right to demand takedowns and pursue compensation for unauthorized digital imitations.

• Google Unveils SignGemma for Sign Language AI Google introduces SignGemma, its most advanced open model for translating sign languages — starting with ASL — enabling developers and the Deaf community to build inclusive applications and expand support to more sign languages.

• EU Toolkit Aims to Help Patients Navigate Health Data Sharing The DSL Toolkit 4.0 explains how the European Health Data Space (EHDS) will impact patients, promoting secure data sharing to improve research, early disease detection, and healthcare innovation — while stressing the need for transparency and public trust in protecting personal health data.

Startups, Tools & Venture Capital

• Databricks Co-founder Launches Laude Institute with $275M Bet on AI’s Future Fast Company covers the launch of the Laude Institute by Databricks co-founder Andy Konwinski, a new organization backed by a $275 million endowment to fund long-term, fundamental AI research.

• INSEADer Startup Adya Launches AI Sales & Marketing Agents An INSEAD-alumni-led startup, Adya, has launched a suite of AI agents designed to automate sales and marketing campaigns, from lead research and personalized outreach to content generation and scheduling.

• NOW for Women: A European FemTech AI Startup Seeks Co-Founding MD INSEADerFounder Natalia Gaydukova is building an AI-powered wellness companion for women 40+ and is actively recruiting a Medical Doctor to join the co-founding team for her European FemTech venture.

• Azure AI Foundry: A Deep Dive into Enterprise AI Platforms A link to Microsoft’s Azure AI Foundry sparked a detailed discussion comparing enterprise-grade AI development platforms (Azure, AWS Bedrock, GCP) with the agility and cloud-agnostic approach of startups like INSEADer-led Adya.

• CodeCleaner AI: An AI-Powered Coding Mentor Enters Private Beta — (from INSEADer Franck Boullier)A new project, CodeCleaner AI, is launching its beta, offering a tool that acts as an AI mentor to help developers write cleaner, production-ready code.

• The Debate on Workflow Automation: n8n, Make, & Zapier (link to n8n) A conversation on workflow automation tools like n8n, Make, and Zapier highlighted their strengths for rapid integration and their limitations regarding scalability and the learning curve for complex, enterprise-grade workflows.

Learning, Resources & Commentary

• How MIT Can Educate the World for the Era of AI (Dec. 2024) An article from MIT’s Faculty Newsletter outlines a vision for the university to educate millions of learners globally on AI fundamentals with an application-oriented approach.

• The AI Revolution: Book Co-Authored by an INSEAD Alum — incl. INSEADer Arnaud Blandin Community member Arnaud Blandin shared a link to the book “The AI Revolution,” to which he contributed a chapter on AI and Sustainability.

• Free DeepLearning.AI Course on Building with Llama 4 A link to a free short course from DeepLearning.AI that provides hands-on training for developing applications with Meta’s new Llama 4 model.

• Peter Thiel’s Views on AI, Mars, and Immortality A New York Times video offers a deep dive into the worldview of technologist and investor Peter Thiel, whose perspectives on AI and the future are influential in Silicon Valley.

• ChatGPT Loses to Atari Chess In a retro showdown, ChatGPT was surprisingly outplayed by the Atari 2600’s simple chess engine — struggling with board awareness and basic tactics while the 1977 AI held firm and claimed victory.

• The Student Essay Is Dead — And So Is Our Cognitive Growth? In a powerful essay, Brian Klaas warns that AI tools like ChatGPT may be killing off the student essay — not just as an academic exercise, but as a vital process for developing critical thinking and human cognition, risking a generation that never truly learns how to think.

• Viral Cat Dives Put Hailuo 02 on the Map MiniMax’s new AI video model Hailuo 02 is going viral with realistic, physics-based videos of cats doing Olympic dives — offering a fast, fun, and accessible rival to Google’s Veo 3 and capturing the internet’s imagination.

• Geoffrey Hinton Warns AI Could Lead to Human Extinction In a powerful new interview, AI pioneer Geoffrey Hinton outlines six existential risks posed by AI, shares why he left Google, and says there’s a real 20% chance that AI could lead to human extinction — urging urgent global action.

• New Pope Makes AI a Moral Priority The new head of the Catholic Church, Pope Leo, is making AI a core focus of his papacy. At a recent AI conference in Rome, he signaled the Vatican may use its moral influence to push for stronger global AI regulation — drawing close attention from Silicon Valley and policymakers worldwide.

• AI in Football: Smarter Decisions, But Coaches Still Key Premier League clubs lead in using AI for tactics and recruitment, but experts warn it can’t replace human judgment — especially in coach-led systems like Spain and Italy..

The AI school with No Teachers

Top 5 Discussions In the Private INSEAD AI Community:

1. The AI Copyright Gold Rush: Training Data as a Legal Minefield A flurry of landmark court rulings siding with AI labs like Anthropic and Meta over the use of copyrighted books for training has ignited an urgent debate on what “fair use” means, creating a high-stakes legal minefield for any company building or deploying generative models.

2. The Platform Wars: Is the Winning AI Moat the Tech or Just Data Gravity? A deep discussion on enterprise AI platforms revealed that the key competitive advantage may not be a superior technological framework but simply “data gravity” — the challenge of moving massive enterprise datasets, giving incumbents like AWS and Azure a powerful, and perhaps permanent, moat.

3. AI’s “Dementia” Problem: The Business Risk of Catastrophic Forgetting Members debated AI’s fundamental inability to learn new information without potentially corrupting or erasing previous knowledge (“catastrophic forgetting”), highlighting a critical reliability risk for enterprises deploying models expected to evolve and improve over time.

4. The AI Safety Crisis: From Hallucinations to “Sabotage-as-a-Service” The community confronted the growing spectrum of AI safety risks, from simple model hallucinations to sophisticated “agentic misalignment,” where autonomous agents could learn to deceive and actively sabotage company objectives.

5. The Cambrian Explosion of AI Agents: From General Tools to Specialized Workers A wave of INSEAD alumni launching startups for specialized “AI agents” — for sales, coding, and wellness — signals a market shift away from using generic, all-purpose models toward deploying tailored AI to solve specific, high-value business problems.

Fun

What’s really inside planets

More

INSEAD AI on YouTube

INSEADers (profs, students, staff and alums) — Join the INSEAD AI private community here

Subscribe to “AI News You Missed” get in your Linkedin feed