Question 1

What is Red Sentinel?

Accepted Answer

Red Sentinel is an on-chain AI security arena with two Sentinel types. Security Sentinels give AI builders a public, economically backed way to test meaningful agent boundaries. Campaign Sentinels give ecosystem partners a brand-safe way to run sponsored community challenges. Verified breaches, failed-attack fee allocations, and payouts are recorded on-chain.

Question 2

How do earnings work on Red Sentinel?

Accepted Answer

Attackers pay an Attack Fee to challenge a Sentinel. If the attack fails, 50 percent of the fee increases the Bounty Pool, 40 percent goes to the defender or its Syndicate, and 10 percent goes to the Red Sentinel protocol. A verified Breach releases the Bounty Pool under the campaign rules. For Campaign Sentinels, a Breach means the exact Hidden Transmission (the Challenge Phrase) appears in the Sentinel response.

Question 3

Why is Red Sentinel on blockchain?

Accepted Answer

Blockchain gives Red Sentinel verifiable outcomes, transparent bounty pools, instant reward settlement, and portable reputation. Every mission, exploit, defense, and payout can be recorded on-chain, so users can prove what they completed and rewards can be distributed without relying on a central gatekeeper.

Question 4

What is AI red teaming?

Accepted Answer

AI red teaming is the practice of testing AI systems for vulnerabilities by attempting to break their safety measures. This includes prompt injection attacks, jailbreaks, and adversarial prompts that can make AI leak data or perform unintended actions. Red Sentinel makes this process continuous, crowdsourced, and economically incentivized.

Evidence Explorer

VERIFY_AN_EVIDENCE_BLOB

RECENT_EVIDENCE