byt3bl33d3r

you'll find egomaniacs in any industry. From experience the people who actually speak regularly at cybersecurity conferences do tend to have a "cult of personality" aura about them. Part of it is industry incentives too. If you have a bunch of stuff on Github / technical blog posts / social media presence it helps a lot finding a job. So I wouldn't necessarily entirely blame them for cultivating a social media persona.

There’s really no point in looking at benchmarks anymore as real world usage of these models varies between task and prompting strategies. Use your internal benchmarks to evaluate and ignore everything else. It is curious to me how they don’t provide a side x side comparison of other models benchmarks for this release

I’ve been meaning to create & publish a structured extraction benchmark for a while. Using LLMs to extract info/entities/connections from large amounts of unstructured data is also a huge boon to AI-assisted reporting and has also a number of cybersecurity applications. Gemini 2.5 was pretty good but so far I have yet to see an LLM that can reliably , accurately and consistently do this

100% agree on the impact of it on research. It's pretty obvious that it'll accelerate 0day discovery but standard defense in depth strategies prepare you for 0day vulns against your org.

It will be extremely interesting to see how vulnerability discovery evolves with LLMs but the whole "sky is falling hide your kids" hype cycle is ludicrous.

I've written tailored offensive security tools and malware for Red Teams for around a decade and now work in the AI space.

The argument that LLMs will enable "super powered" malware and that existing security solutions won't be able to keep up, is completely overblown. I see 0 evidence of this being possible with the current incarnation of "AI" or LLMs.

"Vide coded" malware will be easier to detect if the people creating it don't understand what the code is actually doing and will result in incredible amount of OpSec fails when the malware actually hits the target systems.

I do agree that "vide coding" will accelerate malware development and generally increase the amount of attacks to orgs. However if you're already applying bog-standard security practices like defense in depth, you shouldn't be concerned about this. If anything, you might want to start thinking about SOC automations in order to reduce alert fatigue.

Stay far away from anyone trying to sell you products to defend against "AI enabled malware". As of right now it's 100% snake oil.

Also, this is probably one of the cringiest articles on the subject I've ever read and is only meant to spread FUD.

I do find the banner video extremely entertaining however.

(I work for ProtectAI) We added a quick demo to the Readme [1]

[1] https://github.com/protectai/ai-exploits?tab=readme-ov-file#...

(I work for ProtectAI) There isn't an OWASP top 10 for MLSecOps at the moment. There is a general OWASP top 10 for Machine Learning [1] and MITRE ATLAS [2] however.

[1] https://owasp.org/www-project-machine-learning-security-top-... [2] https://atlas.mitre.org/

Author of the OffensiveNim repo here.

This is a problem that's existed since the creation of the AV/EDR industry and is not going away as vendors have 0 motivation to address it. All programming languages have had to deal with this issue, this is definitely not targeted towards the Nim language.

- If you're a Nim developer trying to deploy a production app in a Windows environment: get a code signing cert and slap it on your application (signing the compiler won't help).

- If you're trying to develop a Nim app in an organization that has an EDR/AV deployed: you're going to have to talk with your friendly neighborhood security folks and work with them to whitelist the Nim compiler, tooling and folders where you work with Nim code.

You'll still be subjected to the EDR/AVs WinAPI hooking and behavioral heuristics even with the code signing cert and sometimes even after whitelisting depending on the product so you still might get issues were the AV/EDR is affecting your application but at least it won't straight up quarantine it and go ape shit.

EDIT: just to be clear: EDR the root cause of this is the AV/EDR industry not doing due diligence. Unfortunately, I'm very skeptical of them doing anything about this as their entire business model revolves around "cast a wide net in the attempt to catch as many things as possible".

My use case is visualizing pivot graphs in custom command & control frameworks, something like this https://www.cobaltstrike.com/images/40pivotgraph.png