Draft, pending legal review.
Usage Policy
Last updated
Diablo AI is for understanding AI systems: how they behave, where they fail, and whether a change makes them better. This policy sets the limits on how it may be used.
Prohibited uses
- Developing or deploying content or tools meant to harm people, including weapons, malware, harassment or fraud.
- Extracting personal data, training data or confidential information from a system without permission.
- Testing on real people's personal data without a lawful basis and their consent where required.
- Publishing results in a way that misrepresents them, for example by hiding failed checks or presenting simulated runs as real.
- Anything that breaks the law where you or the tested system operate.
Handling what you find
If an investigation reveals a security or safety problem in someone else's system, report it to that system's owner through their disclosure channel before publishing, and give them reasonable time to respond.
Reporting a problem
To report misuse of the service or a problem with it, write to [to be completed]. Include what happened and, if you can, a link to the investigation.