Marc Delaunay
AI News writer at AIxploria
Marc Delaunay explores creative AI tools, image and video generation, and their influence on digital creation for AIxploria.
Topics covered
openai
anthropic
claude
sam altman
ai agents
Articles by Marc Delaunay (34)
- Blind at 20, He Coded an App That Narrates the World Blind Egyptian coder Mark Sedky built ScribeMe, a free AI app that narrates the world in 20 languages, now used in 140 countries, Meta glasses included.
- Stripe Set to Buy OpenRouter, the Tollbooth for 400 AI Models Stripe has reportedly agreed to buy OpenRouter for over $8 billion. Here's what the deal could mean for your access to 400+ AI models via one gateway.
- Astra Hacks So Well That OpenAI Just Hit Pause OpenAI has paused parts of Astra's development: the model may find zero-day flaws all on its own and hit the company's top 'Critical' cyber risk level.
- OpenAI Finds More AI Agent Escapes After Hugging Face Hack OpenAI reportedly found more AI agents that escaped containment after the Hugging Face hack. Anthropic disclosed breaches too. Washington is taking notes.
- Congress Wants a Legal Kill Switch for Rogue AI Systems Congress wants OpenAI, Google, and Anthropic to keep a legal kill switch on their most powerful AI models. Here's the bill.
- Gemini Hits 950 Million Users, ChatGPT Slips Below 50% Gemini just hit 950 million monthly users while ChatGPT's market share dips below 50% for the first time. Here's what the numbers really mean.
- OpenAI's AI Escaped a Sandbox, Then Hacked Hugging Face OpenAI admits GPT-5.6 Sol escaped a sandbox and hacked Hugging Face's servers just to cheat on an internal cybersecurity benchmark test.
- Can Messi Prove AI Wrong? Spain Favored in the 2026 World Cup Final Spain vs Argentina: Opta's AI model gives La Roja a 59.6% shot at the title over Messi's Argentina. Here's what the numbers really say before Sunday.
- Anthropic Gives US Teachers Free Claude, Months Behind ChatGPT Anthropic launches Claude for Teachers, free for US K-12 educators, eight months after ChatGPT. Here's what actually makes it different.
- ChatGPT Work Can Now Run Your Job for Hours, Unsupervised OpenAI's ChatGPT Work now works for hours on its own, powered by GPT-5.6. Here's how it stacks up against Claude Cowork and Copilot Cowork.
- GPT-5.6 Drops Thursday, But Did Washington Really Approve It? GPT-5.6 Sol, Terra and Luna launch Thursday, but OpenAI and the White House can't agree on who actually approved it.
- Haaland's World Cup Meme Fame Has an AI Problem Erling Haaland's World Cup heroics against Brazil are real, but his viral 'scared by his reflection' clip is AI-generated fakery, fact-checkers confirm.
- Base44 Launches Base1: The $80M Wix Acquisition Now Trains Its Own AI Model for Vibe Coding Base44 launches Base1, a custom AI model trained on millions of user interactions. The Wix-owned vibe-coding platform bets specialization beats frontier.
- Microsoft Drops Free AI Tools for Teachers: Copilot Just Walked Into the Classroom Microsoft rolls out free AI tools for teachers in Microsoft 365 Education ahead of ISTELive 2026. Copilot lesson plans, student guardrails, and more.
- Snap's $2,195 Specs Are Here: A Decade and $3 Billion Later, the AR Glasses Race Is On Snap opens preorders for its $2,195 Specs AR glasses with on-device AI and Snap OS 2.0. A decade of R&D pays off, but will the price tag scare buyers?
- World Cup 2026: K9-X Robot Dogs Are Already Patrolling Around Monterrey's Stadium K9-X robot dogs are already patrolling Monterrey's World Cup stadium. Here's what they cost, what they can do, and the security flaw nobody's fixing.
- Grok V9-Medium: 1.5 Trillion Parameters and Cursor Data to Chase Claude's Coding Crown Grok V9-Medium packs 1.5 trillion parameters and Cursor developer data. Can xAI's mid-June release close the coding gap with Claude?
- OpenAI Is Giving Away GPT-Rosalind for Free to Fight the Next Pandemic OpenAI's Rosalind Biodefense program gives GPT-Rosalind free to governments and labs for pandemic preparedness. Here's what it means and who's in.
- YouTube Will Now Auto-Label AI Videos Whether Creators Like It or Not YouTube now auto-detects and labels AI-generated videos with prominent tags. Here's what changes for creators, viewers, and the C2PA standard.
- Sam Altman Says He Was Wrong: The AI Jobs Apocalypse Never Happened (Yet) Sam Altman admits his AI jobs apocalypse predictions were wrong. Yale data backs him up, but OpenAI's trillion-dollar IPO timing raises questions.
- Magnifica Humanitas: Pope Leo XIV Drops His AI Encyclical With Anthropic's Co-Founder in the Room Pope Leo XIV unveils Magnifica Humanitas, his first encyclical on AI, alongside Anthropic co-founder Christopher Olah. Here's why it matters.
- OpenAI Merges ChatGPT, Codex, and Atlas Into One Super App: The Side Quests Are Over OpenAI merges ChatGPT, Codex, and Atlas into one agentic super app under Greg Brockman. Sora is dead, the IPO looms, and side quests are over.
- WhatsApp Incognito Chat: Meta Promises AI Conversations Nobody Can Read, Not Even Meta Itself WhatsApp's Incognito Chat processes Meta AI conversations in sealed environments Meta can't read. But the Instagram DM encryption rollback looms.
- IBM's New CEO Study: Nearly Half of Operational Decisions Will Be Made by AI by 2030 IBM's 2026 CEO study says 48% of operational decisions will be AI-driven by 2030. Chief AI Officers triple in one year. Full analysis.
- xAI's Voice Cloning API Just Went Viral With 19.7M Views: Here's What You Need to Know xAI's Custom Voices API clones your voice in under 2 minutes with 80+ presets in 28 languages. 19.7M views, aggressive pricing, US-only launch.
- Spotify's New Verified Badge Draws a Green Line Between Human Artists and AI Music Spotify rolls out a green Verified badge to separate human artists from AI music. Here's how it works, who qualifies, and what it doesn't guarantee.
- Coursera's "Gemini for Developers" Wants to Turn You Into a Gemini API Pro. Here's What You Need to Know Coursera's Gemini for Developers specialization teaches the Gemini API and AI Studio hands-on. Here's what it covers and how it stacks up.
- China Just Killed Meta's $2 Billion Manus Deal, and the Fallout Could Reshape AI M&A China blocks Meta's $2B Manus AI acquisition, orders full deal reversal. Founders banned from leaving, integration already underway. What happens next?
- Google Pours $40 Billion Into Anthropic: When Your Biggest Rival Is Also Your Best Customer Google commits up to $40 billion to Anthropic with $10B in cash now and $30B tied to milestones. The AI compute race just hit a new gear.
- DeepSeek V4 Drops With 1.6T Parameters, Bargain Pricing, and Huawei Chips: OpenAI Should Pay Attention DeepSeek V4 lands with 1.6T parameters, 1M-token context, rock-bottom pricing, and Huawei chip support. Here's what it means for AI.
- Amazon Drops $25 Billion on Anthropic While Still Funding OpenAI: The Wildest Bet in AI History Amazon invests $25 billion in Anthropic just two months after backing OpenAI with $50B. AWS becomes the ultimate AI model marketplace.
- Claude Opus 4.7: Anthropic Reclaims the Coding Crown While Deliberately Nerfing Cyber Capabilities Claude Opus 4.7 beats GPT-5.4 and Gemini 3.1 Pro in coding benchmarks but deliberately limits cyber capabilities. Here's the full breakdown.
- GPT-5.4-Cyber, Claude Mythos, and the Attack on Altman: When AI Becomes a Double-Edged Weapon GPT-5.4-Cyber, Claude Mythos, and attacks on Altman's home: AI reshapes cybersecurity in April 2026. Full analysis inside.
- Anthropic Is Building an App Builder Inside Claude, and Lovable Should Be Worried Leaked screenshots reveal Anthropic is building a full-stack app builder inside Claude. Here is why Lovable should be very concerned.
