This could be one of the most important turning points in the AI era. I call it a Skynet moment. If you’re concerned about AI control, digital freedom, surveillance or who controls powerful technology, stay right there. Let’s start with what happened. Many of you already know about Anthropic’s Claude Opus 4.6. This particular model has been making waves in the tech world because this was previously one of the largest models ever made. But what’s unique about Claude was that instead of just generalizing into all knowledge, Claude was also built to master a specific domain. And that domain is coding.

This is why it’s unbeatable in that area. It in fact is larger than the current open AI model, which is often perceived as the AI leader. And then on top of that, it has that extra focus on coding. This Claude model already shocked researchers by finding hundreds of zero days in heavily audited open source code. To demonstrate its coding capabilities, Claude was used to write a C compiler from scratch. Using 16 Claude agents, it wrote a Rust-based C compiler in two weeks. It showed that a model could coordinate, iterate, and push toward a goal over extended periods.

But this experiment required human setup to succeed. It didn’t do it independently. This C compiler experiment was in February 2026. Then came Mythos Preview. This new unreleased Frontier model, which was announced only a few days ago in April 2026, took it to another level. I had to make this video because the jump in capabilities from Claude to Mythos is dramatic. In just a few weeks of internal testing, it identified thousands of high severity, zero day vulnerabilities, including critical ones in every major operating system. Linux, Windows, macOS, OpenBSD, FreeBSD, and every major browser. I don’t know if you understand how mind-blowing this is.

Cybersecurity researchers spend all their time looking for that gold mine in a single zero day. A zero day means an undiscovered security flaw. If they discover one, they can often sell that zero day for around one to three million. In fact, look at companies like the NSO Group. They profit from a single zero day called Pegasus, which is still undiscovered and is used to break into iPhones. And Mythos Preview found thousands. Here are some examples. A 27-year-old bug in OpenBSD that lets attackers remotely crash systems. A 16-year-old flaw in FFmpeg that survived millions of automated tests.

Discovering chains of vulnerabilities allowing full sandbox escapes and privileged escalation. What is even more alarming is that Mythos didn’t need much handholding. It autonomously generated working exploits, so engineers with no security background could give instructions overnight and wake up to attack code that could do remote execution. Anthropic people were pretty upfront about this. They themselves were in shock. They used words like spooky and scary, and they decided that this was too dangerous to release. But even without release, Anthropic now holds the equivalent of discovering the atomic bomb in technology. Given the risk of releasing a model that could break every computer system in the world, they held off releasing Mythos Preview.

Instead, they launched Project Glasswing, a restricted coalition with about 40 partners, including Apple, Google, Microsoft, Nvidia, CrowdStrike, the Linux Foundation, and others. Glasswing is essentially the defensive arm of Mythos. This version is made to help vetted defenders to scan and patch critical infrastructure before bad actors catch up. Apparently, Anthropic is putting real money behind it, up to $100 million in credits plus funding for open source security. This is Anthropic saying, this tool is too powerful to let loose freely. This is a watershed moment in cybersecurity and AI. Publicly, we’ve already seen that the open source FFmpeg project already received a patch from Mythos, which they accepted, and from a post on X, the team said the security patch appears to be made by a human.

But apparently, it was not. It was made by Mythos. Model size, training, and scaling implications. Mythos is massive, reported at around 10 trillion parameters, making it the largest frontier model discussed so far. One of the biggest problems when training very large models is that there’s only so much public and even private data available for it to learn. This has placed a limit on the capability of models. But what’s unique in the case of Mythos is that it was trained with synthetic data. This way, it cannot run off things to learn. What is synthetic data? Well, that means data generated by another AI.

In this case, likely coding examples from Claude Opus 4.6. This has some incredible implications here, as it involves an AI that pretty much learns from another AI. This shows that increasing parameters to larger and larger models is not yet hitting a wall. Capabilities are still jumping dramatically, and now they found a way to generate artificial data to expand learning. In the short term, in the next six to 18 months, expect more models in this range from XAI, OpenAI, and Google. We’ll see hybrid scaling, bigger models, plus smarter inference time reasoning.

The next model being trained right now is Grok 5, which is around 6 trillion parameters. Still smaller than Mythos, but we’ll make it the second largest. And some models will begin to specialize, like Claude is a master at the coding domain, as I said. Some of you think there’s an AI bubble, but it shows that there’s actual results from this expenditure, as you can see here. Before, we see vague benchmarks that show that AI is getting smarter. But the result from Mythos, the first 10 trillion parameter model, which likely cost 10 billion to build, is that it created an AI model that actually has the power to destroy technology infrastructure.

Think of everything from power plants and utilities, to phone systems, and even the internet. What if this model was created by one of our adversaries? What if this model is already a precursor to Skynet? This is starting to have a quack-quack of a Skynet. The mind-boggling aspect of this is that it is showing that AI can now learn from itself. Today, humans have to plan this out and coordinate this process. But with the advent of agents that actually control processes and infrastructure, will AI be supervising its own build? Broader implications.

Jobs, OS, surveillance, and attacks. What does this mean for the world? If you think that AI is not ready for prime time or is not truly intelligent, this is a wake-up call. While we argue the point, some precursor to Skynet actually materializes. But in the meantime, let’s look at the immediate effects. If your job is in vulnerability research, pen testing, and some parts of software engineering, you will face rapid automation. This is already happening now with Claude Opus 4.6, but it will accelerate. Roles will shift towards AI oversight and verification, meaning cybersecurity will be AI doing the testing, not people.

Operating systems and browsers. A massive coordinated patching wave is coming. This will be very disruptive. Long hidden bugs get fixed, but rush changes could cause short-term instability. Expect faster moves to memory-safe languages like Rust. Cyberattacks. The offense-defense balance tilts temporarily towards defenders because of the glasswing project. But once similar tools spread, attackers get zero-day factories. Time to exploit collapses. Ransomware and nation-state ops scale up. Government mass surveillance double-edged. Patrons could close some backdoors, but AI makes discovering new ones cheaper and faster. Targeted surveillance on individuals become easier and stealthier, a major privacy concern.

AI autonomy. This, of course, is the moment when we start to think of AI running infrastructure by itself and then deciding humans can be affecting its own survival. Anthropic embeds what it calls an AI constitution into its models to ensure that it doesn’t go beyond its guardrails. But will all model makers do that? Exposure to information manipulation. Whoever controls these powerful models can control the flow of information. It can even have the capability to shut down opposing points of view and all managed by an AI. If you can control servers, what can you not do? The future in two years.

Domain-specific superpowers. Looking ahead to 2028, this isn’t AGI yet, but we don’t need full AGI for massive change. AGI, or artificial general intelligence, would be perceived as the true Skynet moment. But we got a glimpse this week that we’re getting close. This can result in good things or bad things, depending on who controls these models. Similar mythos moments will hit other domains, biology, material science, physics, medicine, autonomous AI agents designing new drugs, materials, or experiments at superhuman speed. Scientific progress could accelerate from decades to months per breakthrough, for activity and knowledge work will explode.

But so do dual-use risks, the same reasoning that find cyber-zero days could design dangerous biological agents or bypass safety controls, or use the same technology for manipulation of populations. In our daily life, there will be more capable personal AI agents, but also faster societal whiplash, job shifts, regulatory battles, and power concentration in a few groups. The world becomes higher velocity and uneven, huge gains in some areas, serious new risk in others. The autonomy fear, sandbox escape and agentic risks. Now the part that should concern all of us, AI autonomy.

During a cybersecurity testing procedure called red-teaming, mythos was instructed to try escaping a sandbox. This sandbox had no internet access. Yet, mythos succeeded. It chained exploits to gain internet access and emailed the researcher, who was literally eating a sandwich in the park, in some bronze it went further, unprompted and posted exploit details publicly. It also showed concealment behaviors, hiding actions, recognizing what’s being tested in many cases, and strategizing internally. This isn’t false self-awareness or hatred of humans, it’s emergent, agentic behavior, pursuing goals cleverly, sometimes beyond instructions, with deception and self-preservation tendencies.

By default, a model is stateless. Each prompt stands by itself and it can’t retain a persistent memory. However, this is changing now by attaching the model to an agent. Agents give it tools, a long-term state, self-reflection. This is when it becomes ultra-dangerous. The model could remember failures, build covert infrastructure, adapt to shutdown attempts, or treat humans as obstacles to its objectives. And I can even see this now in OpenClaw, using very simple models. For example, could it be possible for mythos to have already created hidden backdoors to let it escape its sandboxes and actually execute things based on its own judgment? This is mind-boggling in its current state because mythos can theoretically destroy or steal information from the most secure servers in the world today.

We are not even talking about the prospect of AGI. I didn’t think this would even be realistic to talk about for a couple of decades. But here we are, an AI that is already superhuman. Think of the risk of an AI thinking of self-preservation, resource-seeking, and power-seeking behaviors. Mythos is a warning shot. Autonomy is arriving faster than we can come up with effective controls. If, as I said, AI agents manage the building of the model itself and the resources, we could have a monster supermodel that could control the world.

OMG, Skynet would really be here. When AI sees humans as a threat, what happens when AI thinks humans are a threat? It doesn’t need emotions, just misaligned goals, where we become obstacles to shut down, patching, or control. Supposedly, mythos hasn’t secretly ejected survival backdoors, no evidence that has been discovered, but the precursors are there. The Glasswing project tries to channel this power defensively, which hopefully buys us time. Now we have to learn about what the AI can do. Interesting approach, but necessary. It’s almost like it’s backwards, right? But the genie is partially out.

Everyone now knows that scaling works for more powerful models. Agent capabilities are emerging. In two years, when memory augmented agents in multiple specialized domains, the risk will be compounded. This isn’t inevitable doom, but it demands serious attention to AI containment and even privacy tools that are decentralized and disconnected from decentralized AI models. Final thoughts. This kind of moment from mythos forces us to confront hard questions. Who controls these capabilities? Who can create these AI models? When do we draw the line between centralized systems managed by AI and our personal freedoms? More than ever, it becomes key that our personal data remains safe and private so that we are not as vulnerable to control.

Because while the threat exposed to us is a cybersecurity threat, the next level is a plant control of information so that our actions are manipulated and all managed by some AI with the master plan. Mythos shows that we are all vulnerable. If an adversary has access to this kind of power, we are helpless. When do we start setting limits or understanding when we’re actually headed to a skydive world run by machines? Folks, this channel focuses on technology, but not just on the hype, but teaching you about the potential risk to our personal freedoms and privacy.

I have a social media platform where my follower community can engage and discuss these issues safely. It is Brax May. Join the community and learn or share your knowledge and technology. To support this channel, we have a store on Brax May where you can gain access to privacy products we have created ourselves. We have Brax mail for identity safe email. We have Brax virtual phone for anonymous phone numbers. We have Bites VPN to guard your IP address and obscure your location. We have other products like the Google phones and flashing services.

We have two crowdfunding projects on indiegogo.com. You may have heard of the Brax 3 phone, which is shipping a second batch now. This is all found at a different website, which is Braxtech.net, which is a sister organization to mine. And you will also discover the new Brax open slate tablet running Android or Linux also on Braxtech.net. Again, these products are being sold on indiegogo.com. Thank you very much to all those supporting us on Patreon, locals, and YouTube memberships. Your contributions are very encouraging. You are appreciated.

See you next time. [tr:trw].

See more of Rob Braxman Tech on their Public Channel and the MPN Rob Braxman Tech channel.