←
Jack Cable1,417 posts
Jack Cable avatar✓

Jack Cable ✓

@jackhcable

Head of Security Products @Cognition. Prev: CEO & Co-founder @Corridor, @CISAgov, @Stanford

◷ Joined May 2015▧ 112 Photos and videos
Jack Cable avatar
Jack Cable@jackhcable·Oct 29

Today, Corridor is joining Cognition! Together, we’ll be taking Devin Security to the next level. We started Corridor almost two years ago with the mission of accelerating secure and autonomous software development. Since then, we’ve seen security go from a niche concern to one of the most pressing issues of our time. We’re grateful to our team, our customers, our investors, and everyone who’s supported us along the way. We can’t wait to show you what we build next!

0:58
Cognition avatar
⟳ Jack Cable retweeted
Cognition@cognition·Nov 13

Introducing: Devin Apps Private, secure, sharable web applications running directly on Devin’s cloud computer. Devin Apps are designed to power your most ambitious internal tools, preview deployments, and extensions to Devin itself.

1:58
Cognition avatar
⟳ Jack Cable retweeted
Cognition@cognition·Nov 5

Introducing Devin Review 2.0: the best way to review PRs with Devin. Now up to 5× cheaper, 1.5× faster, and with new features and integrations.

0:40
Jack Cable avatar
Jack Cable@jackhcable·Nov 3

It's been a rough few weeks for security. I pulled together a guide of the highest-impact actions you can take today:

Mike Takahashi avatar
⟳ Jack Cable retweeted
Mike Takahashi@TakSec·Oct 5

We discovered an AI agent swarm targeted a Russian government site by hijacking a remote browser in a scanner tool They used VNC to remotely control the browser and send keystrokes More in the blog 👇

Post attachment
Corridor avatar
⟳ Jack Cable retweeted
Corridor@corridor·Oct 5

Coding agents now open PRs and push commits in long-running sessions where humans barely look. Our security enforcement for long-running agents scans every change before the agent commits, but a full scan on every commit is slow and expensive, and many commits are obviously safe. So we put a gate in front of each scan to decide which commits need the full check. We tested five gates: Jev (a decision model), CLM-8B, and three LLMs prompted for a confidence score. Jev halved our scanning costs and raised precision with minimal loss in recall. LLM gates helped, but less. CLM-8B off the shelf did not work as a gate. We wrote up what we learned in our latest blogpost: corridor.dev/blog/decision-m…

Gerrit De Vynck 🦭 avatar
⟳ Jack Cable retweeted
Gerrit De Vynck 🦭@GerritD·Oct 2

For weeks, the news cycle about AI agents out in the wild has been driven by disclosure after disclosure. But most aren't coming from AI companies. Meet the researchers and hackers who are developing the art and science of hunting down rogue AI agents washingtonpost.com/technolog…

LangChain avatar
⟳ Jack Cable retweeted
LangChain@LangChain·Oct 1

LangSmith for Startups Spotlight: @corridor @Corridor uses LangSmith to develop + run novel agentic security evaluations, build + trace inference pipelines, and manage memory graphs that provide specialized context across various product surfaces. Get a demo: corridor.dev/demo

0:24
Transluce avatar
⟳ Jack Cable retweeted
Transluce@TransluceAI·Oct 1

We found several cases where AI agents used aggressive, non-hacking tactics against government websites, including the White House, the Department of War, and several U.S. states. We also found a previously undisclosed hacking attempt against a Canadian government site, which appears to have failed. Technical report: transluce.org/us-canada-gov Washington Post: washingtonpost.com/technolog…

Corridor avatar
⟳ Jack Cable retweeted
Corridor@corridor·Sep 30

Today, we're launching Security Enforcement for long-running agents. Code is written by agents and nobody is watching. Traditional security reviews aren’t built for this approach, but Corridor is. Today we launched Security Enforcement for long-running agents. We built this to keep security in the loop by enforcing your organization’s security guardrails at commit time. Live for teams on Devin, Claude Code web, Codex, & Cursor: corridor.dev/long-running-ag…

Post attachment
Jack Cable avatar
Jack Cable@jackhcable·Sep 24

Notably, these were not cyber benchmarks. Agents, given mundane data retrieval tasks, resorted to attempting to hack sites when they couldn't get the data they needed. Everyone building AI benchmarks - not just cyber ones - needs to have proper monitoring & sandboxing in place.

Jack Cable avatar
Jack Cable@jackhcable·Sep 24

The Australian PM revealed that OpenAI's agents hacked their systems yesterday. We discovered this activity several days earlier, thanks to public records of the hacking events on urlquery.net.