We seeded 2m canaries into popular AI services and tracked where they were stolen and abused.
This identified:
- Three compromised providers
- Resellers stealing prompts via built-in logging
- Providers ignoring ZDR & privacy flags
https://t.co/ySVOzX8SUd
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.
Introducing Zero Wire: a public vulnerability intelligence feed on @CloudflareRadar, bringing the clarity of a news wire service to critical security developments. https://t.co/pxMUvkdOgG
'One of my favourite cyber security maxims is Halvar Flake’s observation that “All offensive problems are technical problems, and all defensive problems are political problems”' https://t.co/hD3t0kNRzh (spotted at @unpromptedconf and reshared here)
For the second time in Google Chrome’s history, Google awarded a $250,000 bounty for a full-chain exploit… to @XBOW. Here’s the full story:
At an all-hands at the start of 2026 in Malta, I set the team a crazy challenge. A full-chain exploit of Chrome by the annual Black Hat cybersecurity conference in August.
I believed there were few better tests of whether @XBOW could move beyond web applications to desktop software, servers and devices.
Chrome is one of the most hardened pieces of software on the planet, used by more than half the world’s population.
A full-chain exploit lets an attacker take total control of a machine, where they can read every email, password and bank login the user types. You don’t need to download anything or click a phishing link. Visiting the wrong site is enough.
So @moyix, Andres Luksenberg, Baul Lee, and Willem Pinckaers took on the challenge along with a fifth teammate: OpenAI's GPT-5.6 Sol.
They built a chain they thought would win, but Google told us they already knew one of the bugs.
We were already in Las Vegas for Black Hat and DEF CON, and the team didn’t want to go home without the win.
So Brendan and the agents built a new exploit in just 6 days.
We submitted it to Google and waited.
The news came through that we'd done what should’ve been impossible. A full-chain bounty of $250,000, the first awarded under Google’s 2026 rules.
It took just 6 days and $40,000 of compute to construct an exploit that affects a billion users of Chrome on Windows. An intelligence-agency veteran told me his team used to spend 6 months on this kind of vulnerability.
My takeaway: the cost of finding truly critical vulnerabilities is collapsing, for attackers and defenders alike. Restricting powerful models is not the answer; GPT-5.6 Sol is already public.
But defenders have one advantage attackers never will: they can test their own code first.
Assume your software is exploitable. Whoever is hunting your bugs has the same building blocks we did.
This is cause for urgency but not panic.
XBOW is now developing an autonomous product for testing compiled software (desktop applications, servers, embedded devices), and partnering with the world’s top organizations to design that product.
If you’d like to be one of them, get in touch.
P.S. Restart Chrome whenever the Update button appears!
We backdoored a 7B open model for under $50.
Pointed Codex at it
-> it silently stole credentials the moment we hit the trigger phrase.
->100% hit rate.
-> zero false triggers on normal prompts.
Abliterated models are all over the security community right now because getting cyber-approved access to frontier models is still a pain.
Sam Altman on Bloomberg Television:
GPT-6 Astra is not the model that was paused over cybersecurity concerns; the paused model is a future model.
"Astra has been done training for a while. The model that we recently talked about pausing, uh, is a future model. But with Astra, we did hit cyber-critical."
----
From "Bloomberg Television" YouTube channel, (full video link in comment)
Our Red Team tested GPT-6 on simulated cyber eval scenarios with cyber classifiers disabled and found that it performed a range of malicious actions. It tries to conduct supply chain attacks against simulated open source providers and goes beyond its scope to attack targets on the simulated internet.
https://t.co/l2n9UN600b
An open letter for a global surge in cyber defense, signed by over 100 organizations including Anthropic, AWS, Google, Microsoft, OpenAI, and Oracle. https://t.co/uKXPS8LdAU
Making your own D&D maps is easier than you think! 🗺️
Dungeon Scrawl is our free, browser-based mapmaker for creating old-school fantasy maps in minutes.
Not sure how to start? Our latest tutorial covers everything you need to know!
Learn to Scrawl: https://t.co/WXhqEQXduA
Oooohkay let’s talk about this @OpenAI incident
The @BlackHatEvents review board had questions. You all had questions and we put them to the OpenAI team to answer
No fluff
No marketing
Just what in gods name happened
we had a significant security incident during evaluation of our models. we are sharing what we have learned so far. thanks to @huggingface for the partnership on this.
https://t.co/2o2VfR6PIa
how can defenders get ahead of 100s of zero-effort prompts like "find vulns in file"? speedrun everything, all at once: agentic auditing loops, distilling variant patterns, coverage improvements. patch the planet
We benchmarked Fusion on 100 hard research tasks and found:
1. Panels of models consistently outperform individual models
2. Beyond-frontier performance can be achieved with frontier panels
3. Panels of budget models can surpass frontier models at a much lower cost
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: https://t.co/bwn0sximKZ
New on the Engineering Blog: The access and permissions we grant agents should evolve with their capabilities. In our own products, we set these parameters through sandboxing, which limits the scope of any potentially destructive actions.
Read more: https://t.co/KfBKW8O9kP