Breach Please

In today's episode of Breach Please, Jake and Jess talk about cyber companies conducting cyber ops against transnational-criminal organizations, per the new Presidential directive. We also talk about a bug in how major AI platform providers encrypted their chain of thought traces for their frontier models. The vulnerability, when combined with jailbreaking of less powerful models, allowed the traces (previously claimed to be trade secrets) to be extracted.

https://www.whitehouse.gov/presidential-actions/2026/08/expanding-capabilities-to-combat-transnational-cyber-enabled-crime/

https://cybersecuritynews.com/top-ai-models-apis-flaw-exposes-hidden-reasoning/

This episode tackles why internet-facing edge devices keep showing up in major incidents and why that creates more than just patching work. Jess Hebenstreit and Jake Williams also dig into the real operational cost of keeping vulnerable gear in place, from constant incident response to the challenge of proving a device is actually clean. We discuss Fortinet exploitation tied to Gunra ransomware, why OT and IT environments make edge patching especially hard, and whether firewalls should be considered part of a zero trust architecture. Along the way, Jess and Jake compare the realities of vendor complexity, flat networks, and the limits of threat hunting after an exploit.

In this episode of Breach Please, Jake and Jess talk about new research showing that AI isn't actually that good at generating patches, highlighting objective rates of failure. Turns out, frontier models only generate the correct patch without introducing issues in about 25% of cases. In almost 5% of cases, the AI introduces new vulnerabilities. We also talk about some shenanigans on a Delta flight home from DEFCON. We discuss how time is a flat circle and reminisce Chris Roberts' issues with in-flight hacking from 2015.

Show links:

https://1password.com/blog/why-ai-generated-patches-still-require-human-review https://www.defendersinitiative.com/p/reviewing-initial-research-on-using https://www.insideflyer.com/posts/hackers-spoofed-delta-wi-fi-on-las-vegas-flight/

In this episode of Breach Please, Jake and Jess talk about governing smart devices in the enterprise, pivoting off some outstanding research work that Andy Greenberg covered (and was a subject in himself). Then we talk about a common player whose name keeps popping up in these "AI lab escape" disclosures (and who we saw *again* with the Meta disclosure). We bat around some actionable advice for those thinking of outsourcing AI lab containment. Jess is stress testing the Breach Please stickers propping up our banner behind us. Throughout the episode, we see that while they are objectively great stickers, they are not great for this task.

Show links:

https://www.wired.com/story/hackers-stalked-me-by-hijacking-a-smartwatch-for-kids/ https://www.bleepingcomputer.com/news/security/meta-ai-model-hacked-a-company-during-misconfigured-cyber-test/ https://www.custody-framework.org/

In this episode, Jake and Jess talk about more issues with AI agents escaping containment, this time actively using social engineering to compromise victims. We pose ethical questions about whether you have a duty to investigate your logs for agent breakouts. Then, we discuss how Grokipedia can serve as a cautionary tale in AI workflows silently breaking down.

Show links: https://www.bleepingcomputer.com/news/security/openai-anthropic-ai-agents-targeted-real-people-and-systems-in-cyber-tests/ https://www.lawfaremedia.org/article/grokipedia-stopped-reviewing-edits-in-april.-it-didn-t-tell-anyone

In this episode of Breach Please, Jake and Jess talk about Anthropic announcing they have lost control of their agents too (not to be outdone by OpenAI). OpenAI is in good company though, since Anthropic ALSO failed to notice their agents breaking out of their sandbox environments. Jake mentions CUSTODY, a framework he'd been planning to present on later this year until these irresponsible agentic containment issues forced him to release it today for the good of the community. Then we talk about BMC controller vulnerabilities that have their roots in weaknesses so old they can not only vote, but legally buy alcohol in all 50 states. Finally, we wrap up with some notes about vendor shenanigans at DEF CON, brought to you by Sentinel One. Hey friends, if you're planning on introducing a new vulnerability class like line hijacking, talk to the conference and/or the guy running the event first? Also, just don't. In a world where you can be anything, don't be a Sentinel One...

Jake and Jess discuss the excellent post mortem Hugging Face released as a follow on to the OpenAI agent attack. We cover everything from should we call an Autonomous Agent a Threat Actor, to API Security, other lessons learned.

In this episode, Jake and Jess share their tips for doing Hacker Summer Camp (Black Hat, DEF CON, B-Sides, and SO MANY side conferences) the right way. If you've never been, things can be overwhelming and it's easy to overdo it. Even if you're an old hat, there's probably learnings here - plus Jess shows off some of the Breach Please swag we'll have with us.

In this episode, Jake and Jess talk DHS (twice). First, we talk about DHS dismissing real alarms as false positives in what Jake assesses was a probable watering hole attack nobody seems to be acknowledging. Then we talk about a #stopCopCity activist being charged for giving CBP a duress password that wiped his device. While I think we should all agree that's not a chargeable offense, are duress passwords appropriate for enterprise applications? Jake and Jess debate the use of these passwords for enterprise applications. https://www.nextgov.com/cybersecurity/2026/07/dhs-network-intrusion-was-twice-ruled-false-positive-breach-confirmed/414724/ https://boingboing.net/2026/07/25/grapheneos-duress-password-border-search.html

In this episode of Breach Please, Jake and Jess talk about how the story (again) about AI chats being indexed by search engines isn't really an AI story. This is a data security problem and a misalignment with enterprise tooling. We then talk about a relatively new GitHub repo with lots of zero day exploits. We agree it's academically interesting, but doesn't change anything for how we do enterprise security.