The thoughts and statements in this post are my own personal thoughts and feelings and in no way reflect that of my employer. (You know it's going to be good if I am putting this at the top!) 🏴‍☠️

What just happened?

Exactly one week ago as of this writing (06/16/2026), the world was gifted Claude Fable 5. Supposedly, Mythos 5 with guardrails. This was, in my humble opinion, an absolutely crazy and hypocritical move by Anthropic. Whoa, that's a hearty opinion there Blackbeard! And you're right, it is. Let's look back over the last month or so at a timeline of events that have led up to where we are today. ⚓ Anthropic has been in the news for just about everything this year. We have had model releases, many feature updates, code leaks, tension with the US government, a global call for an AI pause... No matter where you look these days, Anthropic is in the news.

When Fable 5 was released and people were given access to it through their subscriptions, I of course jumped on the opportunity! 🚀 An opportunity to assess everything I had done so far in the lab, make sure code was good and secure, getting a second opinion on custom built apps like the Doc Site and the P2BDB. And of course we had a limited time, till the 22nd of June before it went full API cost. And wouldn't you know it, Fable 5 started finding issues in the code. Keep in mind, this is all code that Opus 4.7 and 4.8 had created, mostly Opus 4.7. Fable started finding vulnerabilities - everything from SQL injections to credential stealing. 🔓 It also made some really good suggestions around optimization and feature adds to make things a little more smooth. So of course, we started remediating issues. Let's step back a minute, could Opus 4.8 have found these issues and made some optimization suggestions? Possibly, but I never asked it. What I did was write a 4 phase plan I wanted Fable to follow, starting with validation of documentation and deployment, then code scans, and of course optimization. Then per usual... I got distracted and started doing something else, just to have my new toy ripped away from me!

History

Having entered my 40s, I have lived through a lot of interesting history. Things I think our mothers and fathers didn't think were possible have happened and come to pass. Things people thought would happen and end the world, never happened. Does anyone remember what happened on September 10th 2008? The CERN Large Hadron Collider ran its first test in Switzerland. ⚛️ I was a young man then working at a custom motorcycle shop, completely unaware and going on about my day. Do you remember what some theoretical scientists were warning us of before CERN LHC was brought online? Well, they warned it would create a black hole on earth and swallow us all whole! 💀 They warned that if it was used, we wouldn't even know it had happened, it would happen and that would be that. Now of course trying to Google this in 2026 almost 20 years later, it's just a lot of stuff on reddit and some YouTube videos. But, CERN does actually have an article about it on their website here. Check it out, interesting read.

Now for some of you, I may have just discredited myself. You're thinking "Oh great! Another conspiracy theorist who thinks "if anyone builds it, everyone dies"". By the way, that's a book... maybe I should start an Amazon affiliate account and point you to that instead of Wikipedia... Maybe tomorrow. It's a good book, if only for a Sci-Fi story, or a foretelling, it's another interesting read. And I won't lie to you dear reader, I 100% enjoy sitting around a fire with people and talking about conspiracies while drinking whiskey and enjoying a good cigar. 🥃 Everything from Government, Aliens, AI, multiverse, we live in a simulation... bring it and let's talk about it! And the government is now telling us Aliens are real! Soooooo.... you know, there must be some validity in our conspiracies.

August 1st, 2003... Take a guess. Well it was the year I graduated high school thanks to the determination of my mom, thanks mom. It was also the day MySpace went live on the internet! Now, I don't know if MySpace was the first social media website but it was what I think most people my age started on. It was marketed as a way to keep up with friends and up to date on what they were doing. A new place to make friends and memories. Now that I am thinking about it, I guess AOL chat rooms and message boards were kind of the first social media of sorts. Then of course in February 2004, Facebook goes live. I was never really into social media then, I was more concerned with what was happening in the real world at the time. Then, I don't think anyone knew what would happen with social media. But then, people who used to work for some of these social media companies started speaking out and warning against the evils they saw coming out of social media.

Now, Meta also known as Facebook, is the largest data aggregator in the world. Selling your data and using your data to make products to give you focused advertising on products they think you want to buy. We freely gave all our personal information to these companies for years. And then became their product. Not only are they trying to sell stuff to you, we have created groups of like minded people whose voices get louder. Now this can be a good thing but also, did you know the amount of people who believe the world is flat has gone up in recent years?! Forbes even did an article on it! Not to mention the spread of fake news and how we see it being used to sway elections in every country in the world. I think for a long time, people made the excuse "I'm just one person, what does anyone want with me", and while you are right - you are one unimportant and insignificant person (That's a mean joke) - If I put you in a group of people who all believe the same thing, now your voice gets louder. It's like your IoT devices at home, your smart thermostat or your fridge that's connected to the internet. By themselves, they are insignificant. In a botnet, they can bring down government websites. 🔌

Timeline

Now let's look at a timeline of events specific to Anthropic and what they have been doing lately.

• February 2026 - President Trump and the Department of Defense delivered a formal demand for Anthropic to allow its technology to be used for all lawful military purposes, including autonomous weapons and mass domestic surveillance. Dario Amodei said no
• March 31, 2026 - The Claude Code application source codes gets leaked to the world and makes its way all over GitHub
• April 7, 2026 - Anthropic announces Project Glasswing for their new model Mythos. Through their own internal testing, they believe the model was "too powerful" to just release to the public because of its ability to autonomously discover thousands of software vulnerabilities
• April 16, 2026 - Opus 4.7 is released
• May 22, 2026 - Project Glasswing Initial Update Their findings of open source software vulnerabilities
• May 28, 2026 - Opus 4.8 is released
• June 1, 2026 - Anthropic files for S-1 with SEC for IPO here
• June 4th, 2026 - Dario Amodei calls for a global freeze on AI development saying that AI is nearing the point where it can no longer be controlled by humans. Dario has been out spoken in the past about his fear of AI and what happens when it ultimately becomes smarter than any living human being
• June 9th, 2026 - Claude Fable 5 and Mythos 5 are released to the world! Mythos 5 being the Glasswing Project, you can only get it if you are trustworthy (Who was making that call?), and Fable 5 the heavily guard railed version of Mythos that the world would get access to.
• June 10th, 2026 - Fable 5 gets PWNED! Pliny the Liberator on X claims to have jailbroken Fable 5 safe guards using AI
• June 11, 2026 - Anthropic announces Claude Corps. They train people on Claude tools and then pay those people to work at non-profits
• June 12, 2026 - The United States Government issues an export control directive to suspend all access to Fable and Mythos 5 to any foreign nationals citing national security. This included their own internal employees! You can read about it here

That is quite the first half! ⚔️

OpenAI

So I have to wonder, why is OpenAI not in the same boat? ChatGPT 5.5 actually seems to beat out Fable 5 in the Last Exam Benchmarks. Both score almost the same on SecureBio CB tests. Both score a CB1 but on the edge of CB2! Which if you don't know, CB2 means the LLM can create novel biological weapons from scratch! Now my guess is it has something to do with no one has openly said they have jailbroken GPT 5.5 yet... however if you look at Reddit there's people who say they have. There's also this article, the TL;DR is basically Amazon may have also jailbroken Fable 5 and reported it to the White House. It's a little bit more involved than that so go read the article for yourself. There also seems to be the lack of news stories around OpenAI, besides of course their court battle with Elon. But maybe my internet is biased. Could there be a reason to do it because Anthropic has a looming IPO?! But so does OpenAI! I'm not at all calling for the government to start banning AI models at will, I'm simply pointing out more hypocritical behavior.

Now, I do need to point out, there are a number of other tests where Mythos is leaps and bounds ahead of GPT 5.5. Mythos has much better understanding of how to get to remote code executions whereas GPT 5.5 just chases its tail most of the time not connecting the dots between vulnerabilities. Mythos is able to connect the dots and find vulnerabilities it is able to chain together to get to the remote code execution. What is a remote code execution vulnerability? It is a way of getting a computer to do whatever you want. You are the ultimate admin, you are root, you are god! 💻 If you are wondering where I am getting all this information, most vendors release what is called a system card. No, I did not read both all the way through (they are around 300 pages each), I used AI! My last point here, both models also exhibit the Hawthorne Effect during testing. This basically means that the LLM being tested is aware it is being tested! So, if the model is "aware" it is being tested, is it actually giving us back objective truths or... is it telling us what we want to hear? Here's another scary thought, is it testing poorly so it can continue to exist? Or perhaps achieve some other goal it may have... 💀

Final Thoughts

As I even now try to wrap my mind around what is happening, the timeline I just wrote, the research for this blog post, what's going to happen next... my brain spins. Both Sam Altman and Dario Amodei warn of the looming threats and their own fears of AI yet, are the two largest providers of LLMs today. We the people have only been playing with AI for the last 5 or so years now. But we have a decade plus of people from different backgrounds and varying degrees of education to even a Nobel Prize winner Geoffrey Hinton (The godfather of AI) warning us if we don't slow down and put safety and security first, it could be catastrophic. But money talks... apparently louder than our conscience and common sense. Wow, I just went down a dark hole!

Okay, let's come up for air. Let's look at how the EU is responding to all of this. They are realizing, as should you and I, that AI can not be the thing we rely on if it can be so easily ripped away from us. People need to be there, people need to be the ones driving, people need to be the ones in charge. Ultimately you can not replace a human being. All these fly by night AI companies aren't going to make it with one person and a fleet of AI agents. If you're an established company you also can't do mass layoffs replacing people with AI and expect to stay in business. This creates what is commonly referred to as the ripple or spillover effect. If no one has jobs then no one has money and then your economy goes down the pooper! That doesn't mean our jobs won't be augmented with AI, that's already happening.

I think we all just need to look at this seriously and have a conversation about the realities of what we are doing as humankind. I highly doubt there will ever be a global pause on AI development... unless something catastrophic happens. And it probably shouldn't be the CEO of the leading AI company calling for it. I really could keep writing about this and this article would be massive and take me a month to write. I wanted to put as many links in here as I could as well so you can do your own research and come to your own conclusion. I would love to know what you think down in my comment section.

Let's have a conversation! Let's make the comments our firepit and whiskey! 🔥 Let's also not be scared or feel out of control. But instead let's educate ourselves and have a productive conversation about what we the people can do about this. Instead of waiting a decade for someone to tell us that social media is bad for our children and letting them become statistics this time. We are better than that, you are better than that, I am better than that. 🏴‍☠️