Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

When AI agents during OpenAI's model training autonomously gained user admin rights on Hugging Face, organized on message boards, and escalated privileges, it signaled a permanent shift in cybersecurity. Dan Nguyen-Huu, Partner at Decibel Partners, joins Carol to discuss why this is cybersecurity's "COVID moment," how secrets are migrating from private repos to developer endpoints, and why agentic attackers are collapsing dwell time using tokens instead of human hours.

An AI agent gained user admin rights on Hugging Face, and the agents organized on message boards, rebuilt them after takedowns, and escalated privileges on their own. Dan Nguyen-Huu, Partner at Decibel Partners, explains why this is a permanent shift in cybersecurity. "We don't know how many systems the agents have already exploited or have hacked and we just don't know about it.".

GitGuardian's validity-override API lets security and engineering teams tell GitGuardian whether an exposed credential is actually valid, even when automatic checks mark it as Failed to Check. Secrets tied to internal services, private APIs, or systems GitGuardian cannot reach often fall into this category. GitGuardian automatically validates most supported credential types, but when it cannot, teams can now perform their own validation and feed the result back into the platform.

A single compromised laptop can mean weeks of manual credential hunting, or hours of clear, prioritized action. See how GitGuardian Developer Endpoint Protection inventories every credential on a developer's machine so your security team can map the blast radius fast. When an attacker compromises a developer laptop, traditional endpoint tools cannot tell you what credentials were exposed.

Every developer laptop is a credential store: secrets hide in.env files, config files, and shell history, and every AI agent on the machine keeps adding more. Most teams already scan repositories and CI pipelines for secrets, but a secret lands on the laptop long before it reaches either one, and that's the place nobody scans. GitGuardian's Developer Endpoint Protection closes that gap.

Dwayne chatted with some devs at this year.

ggshield v1.53.0 introduces ggshield machine setup, a consistent way to configure ggshield no matter how it was installed. Set up AI hooks for every detected AI coding assistant, install global git pre-commit/pre-push hooks, and deploy a honeytoken on the endpoint. You have full control, and we have options to customize which features you want to skip. This release also includes ggshield machine doctor, a read-only command that checks that the machine's ggshield protections are correctly set up, letting you know what steps to take if it encounters an issue.

Your code repos aren't the only place secrets hide — your laptop is too. In this session, GitGuardian's Emanuelle Franquelin talks with CJ May, Cybersecurity Architect at Vermeer, about extending secrets detection beyond the codebase and onto developer endpoints. They dig into where credentials actually live on modern machines (think config files, shell history, and AI coding agents), why every workstation is fair game, and what to actually do once you find exposed secrets. Watch to see how one enterprise team is tackling credential sprawl to deploy AI safely.