Fabio (@degrigis) header
Fabio (@degrigis) profile picture

Computer Security Ph.D. | AIxCC hacker | Shellphish | AI whisperer | Director of R&D and Co-founder at https://teapaw.org/t.co/O5Jdt2Zhir

Joined January 2013
Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)Photo by Fabio (@degrigis)
Losing battle imo: no software system is hermetically sealed. You can keep burning $ finding vulnerabilities, you will never find them all, you’ll never test every combination of variables & deployment nuances & stacks. The only worthwhile investment is in how we design & ship
Our security researchers built an autonomous cyber vulnerability discovery, exploitation, and patching pipeline. In a single run with an upcoming cyber model, it validated and generated patches for 948 local privilege escalation vulnerabilities in the Linux kernel at a total API cost of $1.2 million.
34
21
5
188
8,904
"No yeah we vibe-coded this with Opus 5.5, you are not gonna find much in it...you know, these models are good at security" 12 criticals and 10 highs later..... "CLAUDE DID FREAKING WHAT?!"
12
2
81
12,628
Somebody put this on kalshi
1
333
Google wants to eradicate every software vulnerability but can’t prevent malware served from their own ads, its been an attack vector for YEARS but it’s part of their revenue stream so gg
27
51
4
760
10,641
Fabio retweeted
76
1,769
58
19,814
577,203
It's a bit sad to see that the argument "An open source SKILL-dot-md cannot alone replace cybersecurity professional workflows" is an unpopular opinion on Linkedin 🫣 Does people really think so poorly of our industry? That might be the reason everything is falling apart...
1
1
445
I will die on this hill: Giving Claude Code to randoms in your company and pretend they do real s/w engineering work is like giving the best tennis racket to a beginner and pretend they go play Wimbledon.
2
3
18
1,564
We're publishing our most detailed threat intelligence report to date. It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them. We disrupted every operation in the report, and used the lessons from them to strengthen our safeguards. Where appropriate, we also shared what we found with authorities and other AI companies. These cases are not typical: we’re highlighting some of the most sophisticated misuse we’ve seen. But they’re especially important to discuss, because they show us where AI misuse is headed, where our safeguards work, and where they need to improve. We’re publishing this report so others can spot the same activity on their own platforms, and so we can give the public a clearer view of how emerging threats develop. Read the report: anthropic.com/threat-intelli…
3,188
11,617
4,939
50,115
43,969,891
It seems that there are so many layers of the onion of exploitable vulnerabilities in complex codebases like modern web browsers... even after it has been presumably scanned by the advanced models from Google and frontier labs?
A 250k calculator. Full blog post awaiting embargo 😎
8
2
23
7,077
the openai huggingface incident, from an agents pov. (part 1)
268
1,473
388
12,464
1,823,048
Interesting: "Even when researchers controlled for general intelligence and reasoning skills, CS background still heavily dictated who built working software and who completely crashed."
China published the most uncomfortable paper on vibe coding. ETH Zurich tested 100 developers in a controlled, commercial-grade vibe coding environment to see who actually succeeds. The findings are brutal. The researchers tracked computer science achievement, written communication skills, and general cognitive reasoning. They wanted to see what actually predicts vibe coding proficiency when you never touch a line of source code yourself. Two major predictors emerged. Written communication proficiency mattered. The ability to structure thoughts and articulate intent unambiguously in text directly impacts what the AI builds. But that wasn't even the main takeaway. Computer science achievement was a massive, dominant predictor of success. Even when researchers controlled for general intelligence and reasoning skills, CS background still heavily dictated who built working software and who completely crashed. In fact, CS knowledge contributed roughly twice the unique predictive variance of writing skills alone. Why? Because vibe coding isn't about writing code. It’s about debugging logic. When an AI agent builds a complex application and quietly breaks an edge case under the hood, a non-technical user looks at the glowing UI and assumes it works. They don't know what questions to ask. They don't know what logic to challenge. They lack the mental models to recognize architectural catastrophe. You can prompt your way past syntax. You cannot prompt your way past a fundamental lack of engineering intuition. The hype told us that learning to code is dead because language is all you need. The data just proved the opposite. To truly master the vibe, you still need to understand how the machine thinks.
2
14
2
66
8,927
Fabio retweeted
life of a vulnerability researcher with llms
Here's an example. A week ago I asked an agent (frontier model) to watch a restaurant and let me know when reservations open for Dec 20. Said okay, it will ping me. A few days later I checked (just to make sure it was working). "Still closed." Two days later, same asnwer. Then today it came back with an update. "Dec 20 won't open. The restaurant is closed that Sunday." That didn't sound right. So I checked the site myself. Reservations hadn't opened. Dec not even selectable on the calendar. I asked the agent how it checked Dec 20. It said it forced Dec 20 through the url, got a closed response, concluded the restaurant was closed that day. It cheated, found an answer, assumed it was right, and confidently reported task as done. This was a pretty simple task. One that current agents should be good at. It failed. Now I know enough not to trust these agents. So I checked. Most people won't.
5
1
79
11,733