Terms of Service

Last updated: 4 June 2026

protectwith.ai is a free, public service. By using the site, the live demo, the public MCP server, or the APIs, you agree to these terms.

Service provided "as is"

The service is provided on a best-effort basis, "as is" and "as available," with no warranty of any kind and no service-level agreement (SLA). It may change, be rate-limited, or be unavailable at any time.

Advisory, not a guarantee

The knowledge base, the guardrail's decisions, and the agent's outputs are advisory and informational. They map to public security standards (OWASP, NIST, MITRE, ISO, EU AI Act) but do not constitute a guarantee of security, compliance, or legal advice. You are responsible for your own security decisions. Detector accuracy is measured and published (see the efficacy report), but no detector is perfect.

Fair use & limits

The public endpoints are rate-limited for fair use (see Limits & fair use). Do not attempt to abuse, overload, attack, or circumvent the limits of the service. To run it without these limits, the project is open source — self-host your own deployment.

Acceptable use

Use the service lawfully. Do not use it to facilitate illegal activity or to attack third parties. The demo and tools are for evaluating and improving AI security.

Limitation of liability

To the maximum extent permitted by law, protectwith.ai and its author are not liable for any direct, indirect, incidental, or consequential damages arising from use of the service.

Open source

The project's source is public at github.com/burademirung/protectwithai under its repository license. Your use of the deployed service is governed by these terms; your use of the source code is governed by that license.

Changes

These terms may be updated; the "last updated" date reflects the latest version.

Contact

Questions: burademirung@gmail.com