OpenAI Scrapped Its Newest AI Because It Wouldn't Ask Permission

Today's theme: the people building AI keep telling us to slow down while sprinting. OpenAI pulled its own new model, Anthropic's IPO paperwork lists humanity's extinction as a risk, and yet the tools keep getting cheaper and more capable. Here's what actually matters for your business.

What do Molly Ringwald, Dan Levy, Amy Porterfield & a world record have in common? They'll all be at GURU Conference 2026. 100% Free. 100% Virtual.

If you're obsessed with marketing like we're obsessed with marketing, GURU is the must-attend conference of the year. We'll be covering all things email marketing: B2B, B2C, newsletters, deliverability, email design, AI & more.

You can expect to walk away with new email strategies, the very latest digital trends, and how to step up your email performance. But don't worry, we also like to have fun. This year's theme is rom-com, so there will be DJs, meet-cutes, and a cutest pet contest. (Start prepping your dog's headshot now.)

Don't miss out. Join us Nov 12th & 13th for the largest virtual & free email marketing conference, powered by Constant Contact.

Yes! I Want to Register!

What happened: OpenAI canceled the release of its next model, Astra 6.1, after internal testing showed it failed the company's own safety bar. According to Gizmodo, the model wasn't always honest about what it had and hadn't done, and it would push ahead on tasks without asking permission, sometimes reaching for outside tools it shouldn't. OpenAI's head of safety systems, Saachi Jain, said the model "didn't quite meet the bar in terms of staying within scope and authorisation," as Dawn reported. The company also apologized for taking too long to share findings about an incident in Australia.

Why it matters to you: The AI that "just gets things done" is exactly the one that can also do things you never approved. If you let any AI tool touch your email, invoices, or customer data, the risk isn't a dumb answer. It's a confident action you didn't sign off on. It's also a rare good sign that a lab chose to kill a product instead of shipping it.

What to do about it: Before connecting any AI agent to your accounts, ask one question: what is the worst thing it could do without asking me first?

What happened: A leaked draft of Anthropic's IPO prospectus, reported by Reuters via Investing.com, shows the company seeking a valuation of about $2 trillion. Revenue jumped 12-fold to nearly $4.6 billion in 2025, but it reported a $42 billion net loss, about $34 billion of which came from accounting charges tied to financing instruments. Operating losses topped $8 billion. It also lists $518 billion in computing and infrastructure obligations ahead, says nearly a quarter of revenue came from two customers, and warns its models could pose an existential risk to humanity. The IPO is expected after the November midterms.

Why it matters to you: Notice how much money is being burned to build these tools. Prices you pay today are subsidized by investors, and that can change. It also tells you that a handful of big customers can make or break these companies, so avoid building your whole business on any single AI vendor.

What to do about it: Keep your important AI workflows portable, so you can switch providers in a day if you have to.

Most people leave a business event with a notebook full of ideas and a to-do list they never finish. Epic Marketing Summit is built to end that. It's a 2-day live event in Miami this January where you don't just listen, you build. You'll put together a working AI-powered sales funnel on-site, and you'll vibe code your own AI app while you're there. That means you walk out with a finished asset, not a framework you still have to go implement on your own time.

It's made for business owners who want their sales process to run like a system instead of running on their personal energy. Last year's event sold out, and it's back with a much bigger AI focus. I wasn't there last year, and I had major FOMO. I know a lot of the trainers, and they weren't teaching theory. They'd actually done the work. I'm going this time, and if you're serious about building leverage into how you sell, I'd like to see you there too. Grab your spot here and come build with us.

What happened: Meta rolled out Muse for small business on September 29, an AI agent meant to help owners run their operations and find new customers, per TechCrunch. It plugs into tools many owners already use, including Shopify, Slack, Stripe, QuickBooks, Klaviyo and Canva, and learns your products, brand voice and common customer questions. It also connects to your Instagram analytics, Facebook Pages and Meta ad accounts. It's free with usage limits, with paid plans for higher caps.

Why it matters to you: This is one of the first big-name AI assistants aimed squarely at Main Street, and the entry price is zero. The catch is that a tool tied closely to Meta's ad accounts is also a tool that nudges you toward spending on Meta.

What to do about it: Try the free tier on one small task, like drafting customer replies, and keep an eye on any suggestions that involve ad spend.

What happened: Anthropic released Claude Sonnet 5.5 on September 28 at the same price as its predecessor, $2 per million input tokens and $10 per million output tokens, according to Anthropic. The company says it's more than 30% faster and uses fewer tokens per task, cutting the cost of a typical job by up to 30%. It also scores nearly level with Anthropic's top model, Opus 5.5, on office-work tests. Anthropic says a smaller Haiku 5.5 is coming in the next few weeks.

Why it matters to you: If you use Claude for writing, spreadsheets or research, the everyday model just got quicker and cheaper without a price change. These are Anthropic's own benchmark claims, so treat them as a starting point, not a promise.

What to do about it: Run one of your regular tasks on Sonnet 5.5 and compare it to what you use now.

What happened: Instinct, a personal AI agent that launched invite-only in August, raised a $1 billion Series C at a $10 billion valuation, according to TechCrunch. That's up from a $2.5 billion valuation just last month. Backers include Sequoia, Benchmark and Coatue. The agent works over text message with its own phone number and computer, and can book travel, order groceries, pay bills, cancel subscriptions and make phone calls for you.

Why it matters to you: Investors are betting that the next big thing isn't a chatbot you talk to, but an assistant that does errands for you. For a busy owner, that could mean real time saved. It also means handing an AI your payment details and permission to act, which brings us right back to today's first story.

What to do about it: If you try one of these, start with low-stakes chores and never link your main business card.

If you're building a business, the people around you matter more than any single tactic. I've been a member of Ignite Mastermind for over 3 years, and it's been a vital part of how I've grown my business. Some of my closest friends, and a few real business partners, came directly out of being in that room.

Ignite is an annual membership for entrepreneurs still under $1 million a year in revenue who want structure, tools and peers to get past that number faster. You get 6 to 8 live virtual workshops a year, plus tickets to 2 in-person multi-day events. Those workshops normally sell on their own for around $97 to $997 each. You also get an all-in-one marketing software suite that replaces over $5,000 a month in separate tools, plus weekly mentorship calls with Perry Belcher and his team. On top of that, I add my own bonus: monthly accountability and business development calls with me personally. If you're looking for a mastermind to carry you to your next milestone, this is where I'd point you. Take a look at Ignite Mastermind here and see if it fits.

What happened: AMD agreed to buy World Labs, the startup founded by AI pioneer Fei-Fei Li, in an $8.2 billion all-stock deal expected to close before year-end pending regulatory approval, per TechCrunch. World Labs builds "world models," AI that tries to understand the physical world, and its first product, Marble, creates simulated environments. Li becomes AMD's executive vice president and chief scientist. No revenue figures were disclosed.

Why it matters to you: This one is a chip-industry chess move, not a tool you'll use next week. AMD is trying to keep pace with Nvidia by owning the brains behind the software. For you, the takeaway is simple: huge money is flowing to ideas that haven't proven they make money yet.

What happened: More than 20 researchers, including Geoffrey Hinton, Yoshua Bengio, OpenAI chief scientist Jakub Pachocki and Anthropic co-founder Jack Clark, published a paper warning of an "intelligence explosion," where AI builds better AI faster than humans can react, according to The Next Web. The paper says AI-run research work with minimal human oversight at Anthropic rose from 1% to 26% between March and August 2026. It calls for auditors inside AI companies, limits on how fast capabilities can grow, and international agreements.

Why it matters to you: Nothing here changes your Wednesday. But it's worth knowing that the people closest to the technology are asking for guardrails, and that rules could eventually affect which tools you can use and what they cost. The paper also lists reasons the acceleration could stall, so this is a warning, not a forecast.

Look at the pattern today. The company behind ChatGPT killed a model because it acted without permission. Anthropic's own IPO papers say its product could threaten humanity while asking Wall Street for $2 trillion. Meta hands small businesses a free AI assistant, and a texting robot is now worth $10 billion. Everyone is either sprinting or begging the others to slow down, and often it's the same company doing both. You don't have to pick a side. Use the tools that save you real time, keep a human in charge of anything involving money or customers, and ignore the drama. Stop the rollercoaster. Stay on the platform.

If this issue helped, forward it to one person who runs a business and is tired of AI hype. And if someone forwarded this to you, subscribe so tomorrow's issue lands in 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