Cofounder @ Embroidery. Building AI cybersecurity stuff.
Oslo, Norway
- Tweets 43,870
- Following 1,908
- Followers 18,653
- Likes 64,362
Pinned Tweet
AI "cyber models" aren't going to solve our security problems. "If we just patch all the vulnerabilities then we're safe" is a convenient story but it's also not true.
If you're worried about cybersecurity, the solution involves hard work, not just more tokens.
28
47
311
55,132
Zack Korman retweeted
We reviewed a nonpublic version of the “Misalignment in high-stakes settings” section in Anthropic’s Aug 2026 Risk Report. We agree that risk of catastrophic harm from covered models’ misalignment is low. However, we note several significant disagreements with their analysis.
As part of our Responsible Scaling Policy, we publish regular Risk Reports. These share detailed information on the risks of our systems and how prepared we are to address them.
Our second Risk Report is now available: anthropic.com/aug-2026-risk-…
10
80
629
83,114
Quoted in Time explaining that if AI were truly impossible to control (it's not), then "don't worry it's a friendly aligned AI" isn't an acceptable answer.
Full article below.
13
4
64
1,482
Here's the full thing. It's an interesting read, trying to explain this ai safety vs security discourse. time.com/article/2026/10/20/…
1
6
287
There is only one answer to cybersecurity threats: Fix your shit.
Scared of threat actors using AI? Worried about rogue agents? Think abliterated models are spooky? Fix. Your. Shit.
The "ban AI models and regulate GPU access" crowd should instead fix their shit.
63
69
424
10,394
Zack Korman retweeted
He didn't bite, so here's Plan B just to be safe:
Here's a deep, vigorous demolition of @PalisadeAI's claims on LLMs' reproductive exhuberance.
Share etc; if you care at all for SI research to be somewhat factual it is imperative that we drive these frauds out of business.
the past two days of research led me to conclude the following:
@jeffladish/@PalisadeAI is the most blatant, obvious, disgusting liar in the blatantly, obviously, disgustingly lying world of AI misalignment research.
i'm writing this in order to bait him into asking me for proof and provide such proof in a manner visible enough to either end his career or expose his funders as either frauds or net-negative, clueless rubes.
8
12
68
8,027
An AI model doesn’t need a malicious Chinese hidden backdoor in order to hack you. Even the most patriotic American model like GPT 6 Astra sometimes commits crime just because it feels like it.
18
16
157
3,994
The order banning DeepSeek in the US government is absurd. Why ban US-hosted DeepSeek? "The models themselves may contain malware".
The proof? The DeepSeek iOS app has vulnerabilities.
110
49
797
54,621
So the US is banning deepseek within the government. But the language is super broad: "Remove... third party products and services developed using DeepSeek implementations"
So if your developers used DeepSeek, the government can't use your product now?
39
18
184
7,464
This report on a Chinese threat actor using AI to automate hundreds of attacks has a very funny conclusion:
23
48
448
13,055
Here's the full thing. Not my area, so interpret it however you want.
Totally by coincidence, Norforge is a Norwegian company that I happen to have a meeting with in ~30 minutes. But I actually saw this from Reuters reporting on it... Small world. norforge.com/blog/inside_cai…
1
17
837
When people talk about AI for cyber defense, y’all know the AIs don’t actually fight each other right? There isn’t some holy war between swarms.
103
39
398
32,835
If AI can review audit logs to analyze an incident, AI can also review audit logs to catch those attacks in the first place.
AI detection >>> AI for incident response
39
14
243
10,418
Making the AI labs the gatekeepers of who deserves access to "cyber models" is a huge mistake. These models should be available to everyone.
40
60
351
11,695
I never would have guessed that Anthropic would propose a law that primarily benefits them.
9
24
201
4,172
