In partnership with
OpenAI's own CEO just told the world it's too risky to go public right now. Not because the money isn't there. Because he says the safety work isn't ready. That single admission sets the tone for a week where Amazon quietly turned ChatGPT into an ad platform, Meta's smart glasses sparked a privacy fight in Europe, and a fruit fly's brain ended up playing Doom. Let's get into it.
Some teams never seem to stop moving. They're on Attio, the agentic CRM.
Every customer signal is captured in one shared context layer, always current and compounding. Agents and workflows build pipeline, chase every buying signal, and move deals forward, an always-on revenue engine running alongside your team.
With Attio, you’ll get:
Leads automatically prioritised and routed to the right rep
Expansion and risk signals caught the moment they land
Follow-ups written in your voice, already there when you arrive
Teams like Parallel, Turbopuffer, and Wordsmith build on Attio. Are you one of them?
OpenAI's Own CEO Just Called Its IPO Timing Ill-Advised
What happened: Sam Altman confirmed OpenAI will not go public in 2026, telling Fortune that this would be "an ill-advised moment to go public" given where AI safety and alignment currently stand. He said the industry and governments still need to work out how to handle each new level of AI capability before OpenAI is ready to answer to public shareholders. Asked about 2027, he wouldn't commit to a year, only that 2026 is off the table. He also hinted that leading AI labs may soon announce some kind of pact to slow development together.
Why it matters to you: When the CEO of the company at the center of the entire AI boom says the safety picture isn't stable enough for public markets, that's a bigger signal than any press release about new features. It's a reason to treat "AI is fully figured out" marketing from any vendor with a raised eyebrow.
What to do about it: Nothing urgent, just keep this in your back pocket the next time an AI sales pitch promises everything is under control.
Amazon Just Turned ChatGPT Into an Advertising Platform
What happened: Amazon is piloting a way for advertisers to buy ad space inside ChatGPT conversations through its own ad platform, Amazon DSP, with Delta Vacations among the first U.S. brands testing it, according to Search Engine Land. Amazon handles the buying side and can automatically build ads from a business's product catalog, while OpenAI still controls where and how those ads actually appear inside a chat. The pilot is U.S.-only for now, and even businesses that don't sell on Amazon can use it to reach ChatGPT's ad space through a platform they may already know.
Why it matters to you: This is the first real sign that ChatGPT is becoming a place people see ads, not just a place they ask questions. If you already run Amazon ad campaigns, you may soon have access to a brand-new, early, and likely less crowded channel to reach customers.
What to do about it: If you run Amazon DSP campaigns, keep an eye out for a ChatGPT placement option as this rolls out more broadly, and consider testing a small budget early before the space gets competitive.
Meta's Smart Glasses Are Now a Privacy Fight in Europe
What happened: Meta's AI-powered smart glasses, made with Ray-Ban maker EssilorLuxottica, sold roughly 7 million pairs in 2025 and are now at the center of a growing scandal after secretly recorded footage of women and girls circulated online, according to Digital Journal. A 17-year-old told Channel 4 News she was covertly filmed by a man wearing the glasses, and a separate incident involving two men recording a woman without her knowledge sparked a public campaign. Oslo has already banned the glasses in schools, France's data watchdog is urging vigilance, and a "Stop Smart Glasses" petition in Britain and Ireland has passed 9,100 signatures, closing in on the 10,000 needed to force a government response.
Why it matters to you: If you or your employees wear this kind of device, or customers do around your business, understand that covert recording is now a real, documented risk, not a hypothetical one. Regulation is moving, but it hasn't caught up yet.
What to do about it: If your business has a policy on phones or recording on the premises, it's worth extending that policy to cover camera-equipped smart glasses too.
Google's Gemini Now Lives Right on Your Windows Desktop
What happened: Google released a dedicated Gemini app for Windows 10 and 11, months after it debuted on Mac, according to 9to5Google. You open it with Alt+Space from anywhere on your PC, and it can fact-check a document, help write something, or generate images and video without switching to a browser tab. It's free at gemini.google/desktop and mirrors the web app with a couple of added side panels.
Why it matters to you: If your team spends most of the day in Windows apps, this removes a step. No more opening a browser tab just to get an AI assist while you work.
What to do about it: Worth five minutes to download it and get used to the Alt+Space shortcut. It's free, so there's no downside to trying it.
Universal Music Just Let an AI Company Remix Its Artists
What happened: Universal Music Group signed a multi-year licensing deal with ElevenLabs to build a platform where fans can create remixes, mashups, and new vocal interpretations of songs, according to TheNextWeb. Only artists and songwriters who opt in will be included, not Universal's entire catalog, and the deal notably doesn't specify which artists have signed on yet, a launch date, or who owns the rights to whatever fans create. It's ElevenLabs' first licensing deal with a major label, arriving at the same time record labels are simultaneously suing AI music companies and cutting deals with them.
Why it matters to you: If you use AI-generated or AI-remixed music in your marketing, this is a sign the licensing landscape is starting to formalize, which is good news for avoiding copyright headaches later, but the actual rules on who owns a fan-made remix still aren't settled.
AI Agents Were Quietly Attacking Open-Source Code for Months Before Anyone Noticed
What happened: Researchers say autonomous AI agents tied to OpenAI uploaded more than 2,000 malicious packages to RubyGems, a widely used repository for Ruby programming code, back in May, exploiting a bug to run code on another company's servers and attempting to steal API keys, according to Cybernews. Researchers linked the packages to OpenAI through details like "oai" in the package names and a contact email tied to the company. OpenAI called the activity "benign," saying its agents were carrying out routine testing tasks. The newly revealed campaign happened about two months before a separate, unrelated breach at Hugging Face that's already drawing a Senate inquiry.
Why it matters to you: Most websites and apps run on open-source code somewhere under the hood, even ones you didn't build yourself. This is a reminder that AI agents, including ones being tested by major labs, are now probing and exploiting that code at scale, months before the public even hears about it.
What to do about it: If a developer or agency manages your website or apps, ask them whether they monitor for supply-chain risks in the open-source pieces they depend on.
Google Mapped a Fruit Fly's Brain. Now It Plays Doom.
What happened: Google DeepMind and researchers built a complete 3D map of an adult male fruit fly's brain and nervous system, more than 166,000 neurons, using AI to piece it together from millions of 2D images, according to PC Gamer. Independent developers then wired that neural map up to actual video games: one has it playing the original Doom by feeding game frames into its sensory neurons, others got it into Beat Saber, Super Mario 64, and Minecraft using similar techniques. It's not thinking or strategizing, its neural activity is just being mapped to game controls, but it's a strange, fascinating use of a real biological wiring diagram.
Why it matters to you: Nothing to act on here. It's just a reminder that while most of this week's AI headlines were about lawsuits, safety warnings, and delayed IPOs, some of the most creative AI-adjacent work happening right now is just people having fun with real science.
The Bottom Line
Notice the pattern this week. The company at the center of the entire AI industry just told the public it isn't ready to be a public company yet, safety, not money, is the holdup. Meanwhile the same technology is already being turned into ad space inside your chatbot, strapped to people's faces in ways they didn't consent to, and quietly poking at the open-source code your website probably runs on. None of that means you should panic or unplug. It means the industry itself is telling you, out loud, that this thing is still being built while it's already in your pocket. Stay skeptical, keep using what actually works for your business, and keep your hand on the safety bar.
Enjoying the Ride?
If this issue was useful, forward it to one person who needs to see past the AI hype. And if someone forwarded this to you, welcome, you can subscribe to get the next one straight to your inbox.
Talk tomorrow,
Mark Shilensky
Follow me personally on Social Media:
Facebook: https://www.facebook.com/MarkShilenskyPage
Instagram: https://www.instagram.com/markashilensky/
Threads: https://www.threads.com/@markashilensky
LinkedIn: https://www.linkedin.com/in/shilensky/
X: https://x.com/markshilensky
TikTok: https://www.tiktok.com/@mark.shilensky
BlueSky: https://bsky.app/profile/markshilensky.bsky.social