Hey there, and welcome to our weekly roundup!
It’s been a big one—OpenAI’s U-turn to Non-profit. The other news is that OpenAI is not just building models anymore; it’s helping shape national AI strategies. Google quietly dropped Gemini 2.5 Pro, and it’s already topping the charts. Microsoft’s making moves around AI security, and Visa + Stripe are going full steam ahead on AI agents and stablecoins. Oh, and ChatGPT just got a GitHub integration that devs will want to check out.
Let’s dive in!
Eearlier this week, I shared AI Startups Ideas by YC. Don’t miss it, these ideas are really valuable:
This Creators’ AI Edition:
Featured Materials 🎟️
News of the week 🌍
Useful tools ⚒️
Weekly Guides 📕
AI Meme of the Week 🤡
AI Tweet of the Week 🐦
(Bonus) Materials 🎁
Featured Materials 🎟️
OpenAI Sticks with Nonprofit Model Amid Musk Lawsuit
OpenAI has announced it will retain nonprofit oversight by transitioning its for-profit arm, OpenAI LLC, into a Public Benefit Corporation (PBC), reversing earlier plans to fully shift into a for-profit structure. The move comes amid backlash and a legal battle with co-founder Elon Musk, who accuses OpenAI of abandoning its original mission to serve humanity.
Board chair Bret Taylor emphasized that the nonprofit will remain in control as the majority shareholder, while CEO Sam Altman noted the decision followed discussions with civic leaders and state attorneys general.
Musk, however, remains critical. His lawyer, Marc Toberoff, dismissed the restructuring as a “transparent dodge” that doesn’t address allegations that OpenAI misused nonprofit resources to benefit private investors, including Altman and partner Microsoft. Microsoft has not yet signed off on the new structure.
A judge recently allowed Musk’s lawsuit to proceed, with a trial scheduled for March 2026.
Fd as a nonprofit in 2015, OpenAI drew criticism after forming a profit-driven subsidiary in 2019 to attract major funding, including billions from Microsoft, raising concerns about mission drift.
While UCLA’s Rose Chan Loui called the new Public Benefit Corporation structure a positive step, she stressed the need for transparency and strong governance. CEO Sam Altman admitted the shift adds complexity but said the mission remains the top priority.
Despite the changes, Elon Musk maintains his lawsuit, arguing that OpenAI still strays from its founding ideals. The legal battle continues, with a trial set for 2026.
OpenAI Expands Its Stargate Project with New Government Partnerships
OpenAI’s not just building AI models anymore—it’s now building nations' AI strategies. This week, the company rolled out “OpenAI for Countries,” a new initiative aimed at helping governments develop local AI infrastructure and tools tailored to their people.
The big idea? Bring the $500B Stargate vision to a global scale. That means in-country data centers, custom versions of ChatGPT that speak the language (literally and culturally), and collaborations to boost sectors like healthcare and education.
OpenAI says it’ll co-fund these projects with participating countries, starting with ten democratically aligned nations. The goal isn’t just smarter tech—it’s also about promoting “U.S.-led AI leadership” and spreading the benefits of AI through a global democratic lens.
It’s an ambitious mix of diplomacy, infrastructure, and AI evangelism. In other words: OpenAI’s playing a long game and it's not just about tech.
Gemini 2.5 Pro Quietly Takes the Crown
Google’s not waiting for the main stage. Ahead of I/O, the company just soft-launched Gemini 2.5 Pro I/O Edition—and it’s already making serious noise on the leaderboards.
In early tests, the model jumped straight to the top of the WebDev Arena, dethroning Claude 3.7 Sonnet with a big lead. Gemini 2.5 Pro is built for developers, with big upgrades in frontend/UI work, code editing, and creating multi-step agent workflows. It even comes with new video understanding skills, letting devs turn clips into interactive learning tools.
Oh, and it’s not just crushing in one area—2.5 Pro now holds the No. 1 spot overall on the LM Arena rankings, beating OpenAI’s latest o3 model.
If December was Google’s under-the-radar SOTA flex, this is more of the same: strong models, shipped fast, with minimal hype. Coding vibes? Fully upgraded.
News of the week 🌍
Microsoft Bans DeepSeek App Over Security Fears
Microsoft President Brad Smith told the Senate this week that employees are banned from using the DeepSeek app due to concerns over data being stored in China and potential propaganda risks.
The app isn’t in Microsoft’s store either, though the company has offered DeepSeek’s open-source model on Azure, after customizing it to remove “harmful side effects.”
While DeepSeek rivals Microsoft’s own Copilot, Smith emphasized the ban is about trust and safety, not just competition.
Honestly, I think it’s a smart move. When there’s a chance that data’s being stored in China and the app might filter or skew information, it’s fair for a company like Microsoft to draw the line. It’s just being careful with data and reputational risk, which is pretty standard these days.
ChatGPT Deep Research Gets GitHub Integration
OpenAI just rolled out a GitHub connector for its deep research tool, letting users analyze real codebases, break down specs, and explore APIs—all within ChatGPT.
It’s in beta for Plus, Pro, and Team users, with more access coming soon. Permissions are respected, and while it’s not flawless, it’s a big step toward making ChatGPT more useful for devs.
Sharing is caring! Refer someone who recently started a learning journey in AI. Make them more productive and earn rewards!
Visa Wants AI Agents to Do Your Shopping
Visa just showed up big at its Global Product Drop event, revealing a wave of AI-driven tools and partnerships designed to reshape how we shop—and who’s doing the shopping.
Is this a new era of marketing? Will we have to start advertising products not to people, but to AI agents instead?
The star of the show is Visa Intelligent Commerce, a new platform that opens Visa’s payments infrastructure to developers building autonomous shopping agents. Think of bots that can browse, compare, and buy stuff for you. Yes, that’s the future Visa’s betting on.
“We see a world where AI makes commerce feel more magical and personal,” said Jack Forestell, Visa’s chief product officer.
The company also announced fresh stablecoin partnerships, updates to its Flex Credential platform, and collabs with big names like OpenAI, Stripe, Microsoft, and Perplexity to bring the vision to life.
Historically, Visa’s AI play was all about fraud prevention. Now, they’re flipping the script—putting AI to work for the shopper, not just the system.
With over 4.8 billion credentials already in play across 150 million merchant locations, Visa wants to be the go-to network powering the next generation of AI-enabled purchases, whether it’s groceries, sneakers, or concert tickets.
The pitch? Visa’s not just a payments company—it’s laying the rails for a whole new kind of commerce.
Stripe fast-tracks AI and stablecoin adoption with new launches.
Stripe unveiled major updates at its Sessions event, reinforcing its mission to grow the internet’s GDP. It launched the first AI foundation model for payments, trained on billions of transactions, boosting fraud detection by 64% for major clients. It also stepped into stablecoins, letting businesses hold and send funds globally using both traditional and crypto rails. Following its Bridge acquisition, Stripe partnered with Ramp to launch stablecoin-powered business cards. Less hype, more infrastructure—Stripe is modernizing how money moves worldwide.
Useful Tools ⚒️
Ciro - Build the perfect prospect list in <5 min with AI
Remio - Local-first AI note taker & personal knowledge hub
Pablo.Design - Unfold your website from your logo.
Explorium MCP- Integrate B2B data to any LLM in minutes
Intercom + Fin AI Agent for Startups - The best AI agent and AI-first customer service platform
Built on a unified AI system, it includes an AI Agent for 24/7 customer support, a Copilot for assisting agents in real time, and an AI Analyst for delivering actionable insights to support leaders. Intercom continuously learns from every customer interaction to deliver faster, smarter, and more personalized service experiences.
Share this post with friends, especially those interested in AI stories!
Weekly Guides 📕
What is MCP? (The next big thing in AI. Learn in less than 20 minutes)
How to Create 3D Animated Stories For Free With Canva AI
How to Create a KDP Book Cover with ChatGPT | AI Cover Guide 2025
Which ChatGPT Model Should You Use? (May 2025 Edition)
Smart AI-Powered LinkedIn DM System
AI Meme Of The Week
AI Tweet Of The Week
Siri, are you alive?
(Bonus) Materials 🏆
Meta.AI Tutorial For Beginners 2025
Introducing Anthropic's AI for Science Program
Suno V4.5 vs V4: Complete Suno AI Tutorial (May 2025 Update)
Making complex text understandable: Minimally-lossy text simplification with Gemini
Mistral Small 3.1 announcement
The look in Altman's eyes tells me he's in a constant state of mental u turns. Are they going to u turn on their dabble in military tech?