Cybersecurity stocks fell Friday, with CrowdStrike (NASDAQ:CRWD) dropping 7%, Palo Alto Networks (NASDAQ:PANW) declining 6%, Zscaler (NASDAQ:ZS) down 4.5%, and Okta (NASDAQ:OKTA), SentinelOne (NYSE:S) and Fortinet (NASDAQ:FTNT) losing 3%. The declines appeared tied to reports that Anthropic’s latest AI model was leaked before launch after descriptions were stored in a publicly-accessible data cache, first reported by Fortune.
Details about the new model, codenamed "Claude Mythos," were reportedly stored in a draft blog available in an unsecured and publicly-searchable data store. The AI startup blamed the leak on "human error" in the configuration of its content management system. The leaked draft introduced a new tier of AI models called "Capybara," which would be larger and more capable than Anthropic’s current top-tier Opus model. According to the leaked document, Capybara achieves higher scores in software coding, academic reasoning, and cybersecurity-related tasks compared to Claude Opus 4.6. The draft described Claude Mythos as "by far the most powerful AI model we’ve ever developed."
An Anthropic spokesperson told Fortune the company is "developing a general purpose model with meaningful advances in reasoning, coding, and cybersecurity" and is "being deliberate about how we release it." The draft blog post reportedly warned that the model is "far ahead of any other AI model in cyber capabilities" and could spark a "wave of models that can exploit vulnerabilities in ways that far outpace the efforts of defenders."
Cybersecurity stocks plunge as Anthropic’s ‘Claude Mythos’ leak sparks AI fear
byu/WickedSensitiveCrew instocks
Posted by WickedSensitiveCrew
11 Comments
A company arguing it shouldn’t be blacklisted because it’s not a security risk ends up having a human error data leak that tanks cybersecurity stocks. That’s hilarious.
Never fear, Ka$h Patel is here!
This makes no sense. The whole SaaS apocalypse makes no sense.
Idk this reads like a publicity stunt?
All stocks are falling…
If anything we should be investing in cybersec. AI will bring a whole new age of data leakage
Um… Feels like this should be obvious, but rarely does correlation equal causation.
That’s especially true here. These two things, the leak and price action of several independent firms while the entire market is tanking, have nothing to do with each other.
Oopsie a leak that makes us look awesome. That’s egg on our face!
The only way AI could disrupt cyber security companies would be to have AI agents essentially required on all PCs and that these AI agents would know when you’re attempting to hack and thus preventing you from doing so. Of course, this sound ridiculous, so it would be unlikely that Cybersecurity companies would be outright replaced because they have network effects from analyzing the data from breaches all over the world.
This makes no sense. Apparently ‘Mythos’ is a new model that’s even more powerful than Opus. It’s going to be way more expensive as well.
If Anthropic controls the model and who uses it, how is it going to a “cybersecurity threat”? If the concern is going to be other players creating models that excel in cyber capabilities, should these companies’ services have a higher demand?
Between this and the iran war, it’s slaughtering me and they are both stupid reasons to be in the red 😐