Cognition launches Devin security remediation program

// 1d agoPRODUCT LAUNCH

Cognition launches Devin security remediation program

Cognition has announced the Devin Security Vulnerability Remediation Program, a six-week structured engagement aimed at helping security teams proactively resolve their vulnerability backlogs. Rather than just identifying issues, the program embeds Cognition engineers alongside Devin, which uses Devin Security Swarm to ingest reports, reproduce vulnerabilities in isolated sandboxes to confirm exploitability, and draft verified patches for human review.

// ANALYSIS

Automated runtime validation and patching is the only viable path to clearing enterprise CVE backlogs, transforming application security from alert generation to remediation.

* Traditional scanners flag thousands of issues, but the lack of engineering capacity makes fixing them a low priority.

* Sandboxed runtime reproduction solves the false positive problem, allowing AI to safely propose patches.

* The hybrid model of embedding Cognition engineers alongside Devin is a necessary bridging step to build enterprise trust in autonomous code modifications.

// TAGS

aisecuritydevincognitiondevsecopsvulnerability-remediationautomation

DISCOVERED

1d ago

2026-07-02

PUBLISHED

1d ago

2026-07-02

RELEVANCE

8/ 10

AUTHOR

cognition

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

NEWS1h ago

Claude Fable 5 redesigns BridgeMind website

The announcement details how Anthropic's Claude Fable 5 model was utilized to completely redesign the bridgemind.ai website. The redesign includes a custom lightning bolt animation that triggers on the "Get Started" CTA, as well as a fully functional, 1:1 interactive replica of the BridgeSpace app built entirely from scratch to showcase the model's advanced frontend generation capabilities.

RESEARCH1h ago

Program-as-Weights compiles specs to LoRA weights

Program-as-Weights (PAW) compiles natural language specifications into task-specific LoRA adapters instead of invoking LLMs at runtime. These adapters are hot-swapped into a frozen local model to enable low-cost, offline execution of fuzzy functions.

UPDATE1h ago

Vercel Sandbox adds FUSE support

Vercel has announced FUSE support for Vercel Sandbox, allowing developers to mount remote filesystems like Amazon S3 or JuiceFS directly inside the ephemeral container. By treating remote storage as a local volume, CLI tools and AI agents can seamlessly read and write data across sandboxes without custom syncing APIs.