AI analysis is entirely optional, but highly recommended for significantly higher precision and recall. All findings are produced by HoundDog.ai's deterministic static analysis engine and remain fully deterministic without AI. The AI layer reviews only the findings already produced by the scanner. It never analyzes source code or influences the detection process.
Without AI
Without AI, all findings are generated by deterministic static analysis. You can manually mark findings as false positives, and the scanner will automatically honor those decisions in future scans. For organizations scanning hundreds of repositories, however, manual review can become time consuming. Scans continue to run entirely within your CI environment on standard CPU infrastructure with negligible impact on pipeline time.
With AI
When AI analysis is enabled, HoundDog.ai automatically reviews the deterministic findings produced by the scanner to:
Mark false positives as Ignored
Adjust severities based on implementation context
Generate detailed privacy analysis, regulatory guidance, and remediation recommendations
Because the AI operates only on the dataflow traces already identified by the scanner, source code is never sent to the AI provider. Scanning remains deterministic and continues to run locally within your CI environment on standard CPU infrastructure with negligible impact on pipeline time.
Organizations connect their preferred AI provider by supplying their own API key, allowing existing DPAs, data residency controls, and contractual terms with that provider to remain in effect.
Supported providers include:
AWS Bedrock
Anthropic
OpenAI
Google Gemini
Microsoft Foundry
The screenshots below show the AI integration settings and an example of configuring an Anthropic model.
The example that follows demonstrates how the AI reviews an existing deterministic dataflow finding, automatically adjusting its severity, identifying false positives where appropriate, and generating detailed privacy analysis without ever receiving the underlying source code.
