Anthropic caused prices for CrowdStrike and Cloudflare to fall by launching the AI tool Claude Code Security to find bugs.

The cyber market reacts to the launch of Claude Code Security from Anthropic

*In response to the announcement of a new cybersecurity tool from Anthropic, shares of companies such as CrowdStrike and Cloudflare fell by about 8%. Investors see Claude Code Security as a serious competitor to traditional vulnerability detection solutions.*

What sets Claude Code Security apart from classic scanners
- No static rules – the neural network does not rely on pre‑built databases.

- It immerses itself in application logic, tracks data flows and relationships between components, reproducing the work of a security specialist.

- This approach allows problems that might go unnoticed with ordinary static analyzers to be identified.

How to start using the tool
1. Repository connection – you need to link Claude Code Security to a project on GitHub.

2. Start scanning – the system will automatically scan the code and detect a wide range of threats, for example:
- lack of user input filtering (SQL injection);
- complex logical errors that allow authentication bypass.

Report of discovered vulnerabilities
- Vulnerabilities are sorted by priority.
- For each issue a detailed explanation in natural language is generated, speeding up analysts’ work.
- Under the error description there is a patch creation feature that allows you to get an AI‑generated fix snippet.

Comparison with similar solutions
- Claude Code Security was released about four months after OpenAI’s Aardvark tool.
- Both products have similar capabilities: they test vulnerabilities in isolated environments and assess the difficulty of exploitation.

Development prospects
Experts believe that Anthropic and OpenAI could integrate their systems directly into development pipelines (CI/CD). This would allow automatically blocking code deployments with security gaps and speeding up product releases without compromising protection.