Skip to content
ClaudeClaude

How Anthropic uses Claude in Cybersecurity

Security teams drown in alerts, jumping between tools and query languages for every investigation. Jackie Bow, Technical Lead on Anthropic's Detection Platform Engineering team, shares how her team used Claude Code to build CLUE—a threat detection platform that automates alert triage and cuts investigations from hours to minutes, powered by Claude Sonnet and Opus models. Check out the full case study to learn more: www.claude.com/blog/how-anthropic-uses-claude-cybersecurity

Jackie Bowhost
May 12, 20263mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Anthropic’s CLUE platform uses Claude to accelerate security investigations fast

  1. Anthropic built CLUE to replace fragmented security investigations that previously required jumping across many tools, databases, and query languages.
  2. CLUE uses Claude (via Claude Code and tool use) to query internal data warehouses and organizational knowledge sources like Slack and code repositories for richer context.
  3. In a sample privilege-escalation scenario, Claude proposes an investigation plan, runs iterative queries, assesses likely compromise signals (e.g., suspicious IP reputation), and produces an investigation summary.
  4. Beyond investigations, Claude Code accelerates internal security engineering by helping new hires understand systems quickly and ship features (like a suppression engine) dramatically faster.
  5. The team frames this shift as moving practitioners toward more research-like work by enabling analysis across massive alert volumes and surfacing what actually warrants human attention.

IDEAS WORTH REMEMBERING

5 ideas

Context is the missing ingredient in many security alerts.

CLUE’s ability to pull from company-specific sources (e.g., Slack discussions and code history) helps analysts judge whether an alert is normal for that environment or truly suspicious.

LLM-driven workflows can turn investigations into guided, iterative plans.

Instead of manually crafting queries across multiple systems, Claude proposes steps, executes tool-backed queries, and refines hypotheses as new evidence arrives.

Automated investigation summaries reduce cognitive load and handoff friction.

By compiling actions taken, findings, and rationale into a final report, CLUE helps analysts and stakeholders quickly understand what happened and what to do next.

Security teams can meaningfully compress engineering timelines with Claude Code.

A feature estimated at 1–2 months (a suppression engine) was built in a week by a new hire, largely because Claude Code could explain how the system is structured and working.

Human attention becomes a scarce resource that the system should protect.

The stated goal is to sift through immense alert and event volumes and only escalate what a person should review, improving signal-to-noise.

WORDS WORTH SAVING

5 quotes

There's no precedent for what we're building, and so securing it is also kind of, you know, a new frontier.

Jackie Bow

Before, investigating a security event would be jumping between five to six different tools, running, you know, three to four different query languages over different databases.

Jackie Bow

CLUE is a detection and response platform that we built with Claude Code.

Jackie Bow

One thing that makes CLUE so powerful is it is connected to our internal system, so it has access via tool use to query our data warehouses, but it also has the ability to query, like, internal knowledge of our company, so Slack messages and our code bases.

Jackie Bow

I'm building the tools that I wish that I had, and I'm actually able to do what I feel like is pushing from being a practitioner to being a researcher, kind of a scientist, because I'm getting to test out a lot of the ways that, you know, I've only imagined us being able to process, you know, this immense amount of data and have more visibility into our systems that previously was just out of reach.

Jackie Bow

CLUE detection and response platformClaude Code for security engineeringTool use over internal data warehousesContext from Slack and codebasesPrivilege escalation and admin-access investigationsAlert triage and human-in-the-loop escalationSuppression engine and onboarding acceleration

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome