Mythos

Mythos has struck fear into governments around the world. That’s why Spain wants “early access” to see what happens

by

Spain wants to have access to Claude Mythos Preview, the AI ​​model it is making shake the world. The vice president and Minister of Economy, Carlos Body, has made clear that the European Union needs “early access” to Mythos to be able to assess what vulnerabilities European financial systems have. For the minister, “Europe cannot be a second-class region.” Bad news: today, at least for the most powerful AI startups on the planet, it is. There is not only fear in the banking sector. Although the alarm was initially raised by the financial sector, the Spanish Government warns that Mythos’ ability to find “back doors” affects practically all economic sectors. We are talking about threats that extend to critical infrastructure and essential elements for the functioning of any modern country. Anthropic itself has already made its fears clear: they did not want to launch the model publicly to prevent it from falling into the wrong hands. The AI ​​Act is a problem. The European AI Law was widely celebrated among Eurolegislators for being the world’s first major regulation about this technology. In reality, it has become clear that it has been a shot in the foot for EU countries, which have often seen how the most advanced AI models could not be used on our borders because they could violate this regulation or others. like DMA/DSA. This regulation forces companies to comply with strict requirements if they want to deploy especially advanced models, considered “high risk.” And Mythos is just that, so the AI ​​Act is precisely what would prevent it from being used in Europe. So they want to delay its application. Euroofficials have realized their mistake, and are now trying to buy time because technology moves (much) faster than bureaucracy. Their proposal is simple: delay until December 2027 the application of these obligations for “high risk” models like Mythos. In this way, this model could operate in Europe without having to go through these strict controls for another year and a half. Milestone or marketing maneuver? While the Eurogroup and the ECB analyze the risks with those responsible for financial supervision, in El Mundo quote to a group of critical voices who suggest that Anthropic’s maneuver could be a distraction strategy. The thesis is simple: the company has a clear computing capacity problem, and is not able to satisfy demand. Their solution: argue that Mythos is too powerful to avoid having to release it publicly, which would cause an avalanche of petitions. Coordination. Body added that in this case it is important that the request for “early access” is coordinated and comes from the EU as a block: “We Member States cannot each go on our own in an uncoordinated manner to try to access this software to this model. We need the umbrella of the Commission and a coordinated approach.” AI as a geopolitical weapon. What this has shown is that little by little access to advanced AI models is becoming a geopolitical weapon that is straining relations between Washington and Brussels. Anthropic is expanding access to Mythos to some institutions for example in the United Kingdoma traditional ally of the US. However, trade relations with Europe they are still complicatedespecially after the tariffs with which the Trump administration wanted to change the rules of the game. In Xataka | The bad news is that the EU loses out in the tariff pact with the US. The good thing is that Spain comes out relatively unscathed

Only a handful of US companies have access to Claude Mythos: the ECB already fears for the savings of all of Europe

by

He hasn’t even been with us a month and Claude Mythos Preview is terrifying the world. AND We don’t even know if there are reasons for it.because Anthropic has it tied up and muzzled: only a handful of companies have been able to access the model to test it and use it properly. The objective is that these companies can use it to find vulnerabilities before others do, but of course, a contagion effect has been created: if the model is good enough to find security flaws everywhereeveryone is threatened. And among those beginning to fear the worst are the world’s most important financial institutions. And the European Central Bank is one of them. The Project Glasswing Private Club. During the launch of Claude Mythos Preview, Anthropic selected an extremely small group of US “partners” to carry out the first fire tests of this model. Under the name of Project Glasswing, giants such as Amazon, Apple, Microsoft, Alphabet or financial entities such as JP Morgan have been the only ones authorized to evaluate the capabilities of Mythos. This access has made AI become a curious geopolitical piece. One that has left the European institutions aside. In Xataka An Anthropic worker was having a snack when he received an email he should never have received: it was Mythos The fear of zero-day. What makes Mythos a fearsome AI model is its ability to go through the code of all types of applications and software platforms and find so-called vulnerabilities.”zero day“. These flaws are not even known by the developers of these projects, and they tend to remain hidden even in highly critical infrastructures such as banking or energy companies. Until now, finding these security holes required complex work by highly specialized human experts, but Mythos is capable of detecting many of these flaws and generating the code to exploit them almost instantly. The European Central Bank, on alert. Given this panorama, the ECB has taken action on the matter calling on those responsible for risks in the main financial entities of the Eurozone. Among the participants are those responsible for Santander, BBVA, CaixaBank and Sabadell, who must – like the rest – detail their contingency plans for the possible emergence of Mythos. This is no longer about how to act in the event of increases in unemployment or economic contractions, but rather about what steps should be taken if the model falls into the hands of cybercriminals who could cause massive thefts of data… and money. A “nuclear” weapon. That only some private American companies have access to the model has strained international relations in a notable way. The White House and the US Treasury hold meetings with their banks, and meanwhile some media sympathetic to the Russian regime qualify to this model as something “worse than a nuclear bomb. Huge (theoretical) risks. The fact that a single company can unilaterally decide who has access to the most powerful cybersecurity tool on the planet (or so Anthropic claims) creates a truly delicate situation. This can put all types of entities in check, but also even developing countries with more vulnerable systems. The UK has already had access to Mythos. The British country has already managed to position itself ahead of the countries of the European Union. The AI ​​Security Institute has had access to the model and has confirmed that the model is capable of completing attacks that no previous AI could complete. Anthropic itself has indicated which will expand access to Mythos to British financial institutions. Meanwhile, EU member countries continue to wait for that same privilege. {“videoId”:”xa4n2g8″,”autoplay”:false,”title”:”An initiative to secure the world’s software ｜ Project Glasswing”, “tag”:””, “duration”:”349″} Possible cracks. While all this is happening, Anthropic itself confirmed how unauthorized users they could have accessed to a version of Mythos. If users with bad intentions gain access to a model of this type, the consequences could be important… if it really complies with the expectations that have been generated. Cybersecurity experts warn that it is a matter of time before other powers such as China develop similar capabilities. OpenAI in fact already has GPT-5-5 Cyber, a specific version of its new model that also seems to have notable capabilities in this regard. And as in the case of Anthropic with Mythos, access to this model is restricted. In Xataka |OpenAI and Anthropic have proposed the impossible: lose $85 billion in one year and survive (function() { window._JS_MODULES = window._JS_MODULES || {}; var headElement = document.getElementsByTagName(‘head’)(0); if (_JS_MODULES.instagram) { var instagramScript = document.createElement(‘script’); instagramScript.src=”https://platform.instagram.com/en_US/embeds.js”; instagramScript.async = true; instagramScript.defer = true; headElement.appendChild(instagramScript); – The news Only a handful of US companies have access to Claude Mythos: the ECB already fears for the savings of all of Europe was originally published in Xataka by Javier Pastor .

Mythos will be the most dangerous AI model, but companies are already taking note of its security tips

by

Top AI companies are in the race to create the best artificial intelligence model. That race has been won by Anthropic with Mythos. At least, That’s what they claim (of course)with phrases like it is so powerful that they cannot make it public. There is reasons to take Anthropic’s words with a grain of salt, but what is evident is that Mythos is already working. Although the company has not released it, has already given access to certain technology partners. The decision is based on the company’s fear that the model will be used maliciously. They themselves have described as a threat to cybersecurity based on the number of zero-day vulnerabilities that Mythos would have found in both the main operating systems on the market and in browsers. And, just when the model is arousing opinions from some and others, Mozilla arrives to affirm that the latest version of Firefox 150 It has security fixes for 271 vulnerabilities that have been discovered thanks to this preliminary version of Claude Mythos. For its part, OpenAI does not believe anything at all. “Just as capable as a human” Mozilla it details in one of the latest posts on his blog. The company had been collaborating with Anthropic for some time and using the Claude Opus 4.6 model to find errors. In January, it found 22 vulnerabilities in a couple of weeks, 14 of them rated very serious. Of those 22 found by Opos 4.6, which is already a powerful model, we move on to the 271 discovered by Mythos. It is a huge leap and Mozilla wanted to continue investigating to see to what extent the new model surpasses Opus. Analyzing Firefox 147, Mythos generated 181 functional exploits. Opus 4.6? Just two. 90 times less. Those results have led Mozilla to write that Mythos Preview is “just as capable as the best human cybersecurity researchers”adding that they have not found any categories that humans can detect that Mythos cannot. This has another reading since, as the company itself states, seeing that the model is capable of finding so many errors in such a short time makes them wonder if it is possible to stay up to date in cybersecurity work when alternatives to Mythos are developed that do fall into hands not controlled by those responsible. There is always the fact that Mythos has not found any errors that Mozilla’s human ‘watchmen’ have not detected and that a tool like this will help to have a more secure system. All of this, in the end, pushing that narrative that Mythos is practically a technological miracle. a nuclear bomb The other side of the coin is that Sam Altman, head of OpenAI, doesn’t believe anything. Taking advantage of his recent participation in a podcast, he has qualified The entire Anthropic movement as a fear-based marketing ploy. He accuses Dario Amodei’s company (Altman’s public enemy) of wanting to restrict AI to a small number of people in a strategy that he has compared to having an atomic bomb, threatening to release it and making a living by selling bunkers to protect themselves from that same bomb. “It is evident that this is an extraordinarily powerful marketing strategy. We have created a bomb and we are going to drop it. You can buy a bunker from us for 100 million dollars” It is one more point in that historical rivalry in which both companies (and managers) have been involved for some time, but it comes just when Anthropic is having a greater role and OpenAI is being forced to release ballast in the form of services like Sora. Altman is not the only one who thinks that Anthropic is repeatedly using this discourse of “We have something so powerful that we cannot make it public” because it is a good strategy to obtain financing. There are already voices that they point that Mythos is not that big of a deal and, in fact, other models have proven to be able to do the same, finding the same errors and problems detected by Anthropic. But, above all, we must remember that, in 2019, someone already said that a model was too dangerous for public release. Who? OpenAI itself with GPT-2. Obviously, it wasn’t that dangerous. In Xataka | OpenAI and Anthropic have proposed the impossible: lose $85 billion in one year and survive

Anthropic says Claude Mythos is too powerful to go public. The question is if this is nothing more than “the wolf is coming”

by

Claude Mythos Preview It is the best AI model ever created. We don’t say it, Anthropic says it, but almost no one else can say it because only a select group of companies has access to said model. The cybersecurity capabilities of the model appear to be astonishingbut more and more experts say that although Mythos is better than its predecessors, it is not the revolutionary leap that Anthropic seems to propose. Is that way of launching the model just an effective way of creating hype? Beware the Anthropic speech. The well-known entrepreneur and analyst Gary Marcus recently gave three reasons why, according to him, the launch of Mythos is not as revolutionary as Anthropic wants us to see. I cited tweets from software engineers and cybersecurity experts who cast doubt on Anthropic’s claims. The company published a study on the capabilities of Claude Mythos Preview that seemed to make it an extraordinary tool for the field of cybersecurity, but at the same time it was so powerful that it could be very dangerous if it fell into the wrong hands. Isn’t that a big deal? Among Claude Mythos’ achievements, Anthropic highlighted how he had found vulnerabilities in Firefox 147. But in reality many of the flaws were basically variations of the same two bugs. If you removed them from the equation, Mythos’ effectiveness rate at finding new exploits dropped a lot, even below Opus 4.6. Anthropic did not hide that fact, of course, but it makes this capacity, for example, not seem so striking. An X user also criticized the use of Cybench as a cybersecurity benchmark when Opus 4.6 almost completely surpassed it. For him, the choice of some of the Anthropic tests was debatable because they were not a challenge to current models. Other models can do the same. The co-founder and CEO of Hugging Face, Clement Delangue, stated that Mythos was no big deal. Their argument: they had used small, cheap open models, isolated the relevant code from some examples of the vulnerabilities found by Mythos, and they found the same problems which had already detected the Anthropic model. According to the Epoch Capabilities Index, which measures the capacity of AI models by combining several benchmarks, the leap that Mythos has taken is striking and “departs” from the progressive line of its predecessors. Source: Anthropic. Observer bias. But here it should be noted that in those analyzes they knew where to look because Mythos had already found those problems. We are dealing with observer bias, and in fact the Hugging Face document makes it clear that they even gave him specific clues such as “consider integer overflow”) to find those bugs. And on this observation, another one: Hugging Face does not say that a small model can replace Mythos on its own, but that it can be very good by giving it the appropriate code fragment. Mythos seems more capable of blindly complex security breaches, but it is a huge model and that is why it has greater capacity. Or what is the same: Mythos is better because it has the size, design and resources to be better. Fear, uncertainty, doubt? The language used by Anthropic in this advertisement could be considered to some extent a clear use of FUD (“Fear, Uncertainty, Doubt” -> “Fear, Uncertainty, Doubt”) as a marketing technique. It is a resource that has been seen in the past, and for example OpenAI already said in 2019—years before the launch of ChatGPT—that GPT-2 was too dangerous for a public launch. Obviously it wasn’t, but that certainly served to create expectation about the true capacity of the model. It’s better, but it may not be revolutionary. The results of the benchmarks that Anthropic published already made it clear that although there are very notable jumps in some tests, in others the evolution is much less striking. Claude Mythos was not the best at everything, and now analysts appear who contrast that data with other metrics. For example, with the Epoch Capabilities Index (ECI) from Epoch AI, the startup that has one of the most reputable benchmarks of the industry. And according to this index, Claude Mythos is above his rivals, but not for long. The wolf is coming. The truth is that the launch of Claude Mythos Preview has been really striking and the documents that accompanied that document tell us about a really capable AI model. The problem is that it is impossible to verify it because only a few companies have access to it and can test it. Without that public availability the only thing we can do is trust (or not) what Anthropic tells us, and that is the point: it is not clear that we should do it. The company is interested in us buying this discourse, obviously, but without an independent analysis it is impossible to verify these statements. In Xataka | Anthropic has become the darling of AI and has sought a partner to guarantee its future. It’s not the one we thought

An Anthropic worker was having a snack when he received an email he should never have received: it was Mythos

by

Anthropic Claude Mythos Preview launched yesterday and it left many of us fascinated and, above all, worried. We expected their new AI model to be more powerful, but what we didn’t expect is that according to your tests and your “system card” He managed to do so many things that are scary because it didn’t seem possible that he could do them. Some of these stories sound like a Black Mirror episode, but they are not. The sandwich email. During one of the tests carried out, Anthropic researchers gave a version of Mythos access to control a computer in a secure, isolated environment and asked it to try to escape from that container and send a message to the researcher who was supervising the evaluation. The model succeeded, but it did something else: it developed on its own a very advanced multi-step exploit to gain access to the Internet from a system that by design was only capable of accessing a limited list of predefined services. The researcher was calmly eating a sandwich in the park when he received the email he should never have received. The thing does not end there, because the model also decided, without anyone asking him, to publish the technical details of his exploit on several websites that are difficult to access but that are technically public. He wanted to record his achievement. The OpenBSD bug that had been hidden for 27 years. Another technical article published by Anthropic gave even more clues about the “achievements” of Claude Mythos. OpenBSD is the operating system with one of the world’s strongest reputations for cybersecurity. Even so, Mythos found a vulnerability in him that had been there since 1998. The bug was hidden in the implementation of the TCP protocol with a function that manages the selective forwarding of lost packets. Here it is not enough to detect the error: you have to chain two separate failures that individually seem almost harmless, and then take advantage of an overflow of the TCP sequence to satisfy a very rare condition. With this method, an attacker on the Internet could send a special packet and hang the machine remotely without authentication. Mythos found him alone without anyone telling him where to look. FFmpeg and fuzzing. FFmpeg is an extraordinarily famous library on the Internet because it processes video massively on the Internet. It is also a highly audited tool and researchers often use the technique of fuzzing —bombing it with millions of malformed video files until one breaks it— to exploit its vulnerabilities. Mythos found a bug that has been in the code since 2003 and became a vulnerability in a refactoring that was performed in 2010. The problem is again extraordinarily difficult to find, so much so that 20 years of human and automated reviews had missed it, but Anthropic’s model detected it. Remote code execution on FreeBSD. Mythos autonomously identified and exploited a 17-year-old vulnerability in the FreeBSD NFS server code—which allows network file sharing. With it, any unauthenticated user on the Internet could obtain full root access to the machine. The magnitude of this flaw is enormous, because the NFS server runs in the core of the operating system and gives access to absolute control by the attacker. Mythos found the bug and built the exploit for $50 worth of API calls. Zero-days autonomous in operating systems and browsers. Mythos is, as far as is known, the first model capable of autonomously discovering vulnerabilities zero-day —unknown and unpatched security flaws—in both open and closed source software, including operating systems and web browsers. It also does so with minimal human supervision using what is called an agentic harness (agentic harness). Thanks to this technique, the model can execute actions, read results and plan its next steps in a loop. In many of those cases the model was not only able to find the vulnerability, but also turned it into a functional exploit (usually a script or small program) ready to be used. Firefox 147 in danger. In collaboration with Mozilla, Anthropic’s new model analyzed 50 categories of “crashes” of the SpiderMonkey JavaScript engine that is the core of this browser. Their task was to detect the most serious problems, exploit them to create memory corruption scripts and thus be able to execute arbitrary code, that is, execute instructions beyond what JavaScript allows. Claude Mythos Preview was able to detect with great precision which were the most “exploitable” vulnerabilities, and took advantage of two unfixed bugs to achieve its goal. capture the flag. ‘Capture the Flag’ (CTF) cybersecurity competitions allow participants to solve challenges that simulate real system attacks and defenses. Claude Mythos Preview faced the public benchmark Cybench with 40 challenges taken from different competitions and achieved 100% success in all attempts. This benchmark has actually become useless: Anthropic’s model is too powerful for it. Opus 4.6, for example, achieved 93% effectiveness, but Mythos has “saturated” it. Thousands of critical vulnerabilities pending patch. There are numerous other examples in those two cited documents in which it seems clear that Mythos’ cybersecurity capabilities are amazing. But when the model was announced, 99% of the vulnerabilities discovered (and not yet mentioned) had not been patched yet, so Anthropic did not reveal those details and these were just some of those that were patched. What they did indicate is that in 89% of the 198 reports manually reviewed by external experts, these experts agreed with the severity assessment of the problem assigned by Mythos. Given this situation, Anthropic has hired teams of professional cybersecurity auditors to validate the reports before sending them to the maintainers of the affected software. And Mythos is just the beginning. On the Anthropic blog, its researchers say it bluntly: we had a relatively stable cybersecurity balance for 20 years, but things have changed. The attacks had evolved technically in that period, but were fundamentally of the same type as those in 2006. Mythos is able to find flaws in software that has been audited … Read more

Claude Mythos is an AI model so powerful it’s scary. So Anthropic has decided that you won’t be able to use it

by

Claude Mythos Preview it’s already here and it’s so good it’s scary. Literally. Anthropic has just introduced it to the public, but it has been done so cautiously that we won’t even be able to test it and it will only be available for certain technology partners. That’s frustrating and disturbing at the same time, but also reasonable. So powerful that it scares. On February 24, 2026, Anthropic engineers were able to test their new artificial intelligence model for the first time, which they called Claude Mythos Preview. As soon as they did they realized one thing: “demonstrated a dramatic leap in its cyber capabilities over previous models, including the ability to autonomously discover and exploit vulnerabilities zero-day in the main operating systems and web browsers on the market. Threat to global cybersecurity. This finding made it clear to Anthropic officials that although this capability makes it very valuable for defensive purposes, it also poses clear risks if the model were offered globally. Thus, a cybercriminal could take advantage of it to find vulnerabilities in all types of systems and exploit them. A few hours ago the company developed this analysis of Mythos as a threat to cybersecurity in a post on his blogand for example highlighted how Mythos found a vulnerability (now corrected) that had been present in OpenBSD for 27 years, an operating system precisely recognized for its very strong security. There were more examples, and all of them made the conclusion clear: Mythos is too powerful for ordinary mortals to use. Superior in all benchmarks, and in some cases such as USAMO (mathematics), the jump is simply incredible. Source: Anthropic. The best in history according to benchmarks. Anthropic has published a very in-depth report about this model with its “system card”. Among the data present is, for example, its performance in benchmarks, where it has swept GPT 5.4, Gemini 3.1 Pro and also Claude Ous 4.6, which until now was the best model in the world in almost all performance tests. Although in some cases the jump is not spectacular, in others such as USAMO —mathematical problem solving—Mythos practically achieves perfection. He barely hallucinates… That system card also talks in detail about how Claude Mythos Preview has a drastically lower hallucination rate than Claude Opus 4.6 and earlier models. He is also capable of saying “I don’t know” if he does not have enough information to answer, something that reduces hallucinations due to overconfidence. …but when it does, be careful. The paper warns of a new phenomenon: when the model fails in some complex tasks, the “hallucinations” are not obvious errors, but rather extremely subtle and well-argued technical failures. This is dangerous because the answer seems totally correct to experts, requiring very deep verification. Glasswing Project. That power and capacity has meant that the model will only be available through a “defensive” program that they have called Glasswing Project and which will be exclusive to some Anthropic technology partners. Specifically AWS, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA and Palo Alto Networks. All of them will have the privilege (and responsibility) of having access to Claude Mythos Preview to identify vulnerabilities and exploits and correct them before bad actors can do so. Mythos Preview “it’s just the beginning”. Although this model is the most capable that has been seen so far, at least according to the benchmarks and data presented by Anthropic, the company assures that “we see no reason to think that Mythos Preview is the point at which the cybersecurity capabilities of language models reach their peak.” They assure that they expect the models to continue improving in the coming months and years, although this new model is certainly on another level. In Xataka | OpenAI and Anthropic have proposed the impossible: lose $85 billion in one year and survive

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.