Ryan McGeehan

Writes "Starting Up Security" @ tweets horror stories @badthingsdaily

[Ryan McGeehan @Magoo]

I did a whole lot of IR last month. Much of it was helping companies affected by compromises at their neocloud providers.

I wrote up the lessons and practical steps teams can take to reduce their exposure and make incident response less painful.

https://t.co/PNvxsYzGUk

[Ryan McGeehan @Magoo]

@dinodaizovi Don't we already have least privilege and privilege escalation?

NGL though I look forward to not having computers constantly spell check "privilege" for me in the future. It should def have a "-edge". So bring on the buzzword

[Ryan McGeehan @Magoo]

Worlds collide. My favorite podcast. Can't wait to listen.

[Calif @calif_io]

Our own @brucedang and @XorNinja went on the Bloomberg Odd Lots podcast to discuss AI, cyber, and php2shell.

Check it out: https://t.co/6ylOEk57ZJ

[nisedo @nisedo]

There was no payout because PHP2Shell came out of Patch the Planet (https://t.co/qiDreq4g9u), a pro bono initiative where we (Trail of Bits, OpenAI, Calif and HackerOne) use AI to audit and secure open source software for free.

If we were paid for every bug from PTP we’d all be driving Lambos by now.

[Ryan McGeehan @Magoo]

I have never been spread so thin on incident response coordination over this past couple of weeks.

[Ryan McGeehan @Magoo]

Other way around, distillation not the vector but the motive to get the keys

[Matt Johansen @mattjay]

Distillation? Led to customers API keys being stolen?

How in the hell?

[Ryan McGeehan @Magoo]

Great example of the problem statement owned by _privacy engineering_. Products (Esp. in social) are "just some cool idea" on the surface. Underneath, they become spaghettified from laws, historic public promises, tradeoffs, and IRL abuse after launch.
https://t.co/GeiwcWh3sW

[Thomas H. Ptacek @tqbf]

Today I'm parting ways with https://t.co/u51sOTFs6x. It's been a privilege. I'm off to tilt at a big old windmill with Kurt, again.

https://t.co/rIfFq2dv6e

[Trail of Bits @trailofbits]

.@1Password's FLAWED report says AI models produce a clean security fix only 26% of the time.

Defenders shouldn't take that number seriously.

• The six vulnerabilities were handpicked because their fixes were complex. Clean-fix rates ran from 3% to 60% depending on the bug, and the report averaged them together.

• Agents set up to fail were counted in the headline figure. Two of 1Password's prompts instructed the agent to apply the wrong fix. Those trials make up 22% of the data. One evaluation mode prevented the agent from compiling or running any code, and it accounts for 36% of the data.

• The report ran two models, GPT-5.5 at medium effort and Opus 4.8 at high. Neither was tested at its highest available setting, so the report says nothing about how more effort or stronger models change the results.

• Several instruction and grading errors further undercut the headline, and are elaborated upon in the attached blog.

We've spent four months submitting hundreds of AI-authored patches to widely adopted open-source projects as part of Patch the Planet.

Our experience didn't match 1Password's report, so we did a full analysis across 186 AI-authored pull requests and 33,500 subsequent commits, benchmarked against 2,265 human-authored patches we graded across years of security engagements. https://t.co/mdgED6y6eB

[Trail of Bits @trailofbits]

1,535 potential bugs found, 1,017 awaiting patches, 326 fixes open upstream, 192 merged.

Since Aug 4: +398 bugs, +46 merged, 55 codebases under review, and scapy has overtaken vllm as the most-reported codebase in Patch the Planet.

https://t.co/dDV7RUbxgQ

[Ryan McGeehan @Magoo]

This talk is a must watch. Think about what agentic intrusions could look like and how incident response will be shaped after them.

[DANΞ @cryps1s]

Our Black Hat talk on the OpenAI-Hugging Face incident is now live on youtube.

This is a watershed moment for the industry. I encourage all defenders to watch, consider how attack dynamics will imminently change, and plan for accelerating defense.

https://t.co/5B6k7YfpcP

[Ryan McGeehan @Magoo]

OpenAI speaking tomorrow about the Hugging Face incident at Blackhat.

[DANΞ @cryps1s]

Black Hat invited us to speak tomorrow about the Hugging Face incident. Given its complexity, we think it’s important to share what happened, what we learned, what we’re changing, and what this means for AI security and alignment. We still plan to publish a technical postmortem once the review is complete.

[Trail of Bits @trailofbits]

858 potential bugs found, 595 awaiting patches, 120 fixes open upstream, 143 merged.

Our new Patch the Planet dashboard breaks down the latest results from our audits of curl, OpenSSL, Kubernetes, and 38 other open-source projects.
https://t.co/iJxPqKLL7M

[Trail of Bits @trailofbits]

Rust maintainers assumed a team of engineers was behind our bug reports. It was one engineer using Codex's /goal.

[Dan Guido @dguido]

Here's the big thing @trailofbits was cooking:
https://t.co/DQ55EbSP3X

[Ent @Ent_Security]

🚀 We're out of stealth.

Today we're introducing Ent, the industry’s first intent-aware Workspace Security platform for human and AI-driven work

🎥 See what it means to protect work as it happens.

#IntentAware #LaunchFromStealth #WorkspaceSecurity #CyberSecurity

[Mitchell Hashimoto @mitchellh]

I strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out.

I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really).

It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely.

The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture.

We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying.

I worry.

[Ryan McGeehan @Magoo]

@crtheisen @bettersafetynet Hi @bettersafetynet - Here's a good starting point to get acclimated if you're looking for one: https://t.co/OIm16jaLr1

Happy to DM if you run into trouble

[Ryan McGeehan @Magoo]

I wish all security pros practiced a scenario-first mindset. Explanations based on risk scenarios before jumping to best practices, gaps, controls, compliance etc. I wrote an essay to coach on this: "Writing a risk scenario"

https://t.co/mvft7zeClo

[Ryan McGeehan @Magoo]

I wrote about that moment every security team faces when someone asks if they can work from China for a while, and then everyone freaks out.
https://t.co/4e8BzwMDdt