RedTeam (SN61)’s Immune System To Turn Unknown Attacks Into Known Defenses

Cybersecurity has always been reactive, with defenders responding after attackers discover new bypasses. Innerworks, the parent company, used Yuma’s Subnet Spotlight to announce the Immune System, launching September on RedTeam (SN61).

The engine generates novel attack vectors, then builds defenses before those threats ever reach the wild. It builds on six months of development and a mining loop ingesting real attack code globally.

The Immune System, Explained

The mechanism sits on top of the existing RedTeam mining loop and extends what miners can produce beyond anything reachable from outside the network.

1. Ingests miner attacks continuously as seed uniqueness: Every submission becomes a starting point for the internal engine to expand from.

2. Accelerates attack vectors beyond external capability: Combines miner code with full internal exposure to Innerworks’ defense stack, generating variations no outside attacker could construct.

3. Predicts attack vectors before they appear commercially: Novel attacks are created in-house first, then reverse-engineered into detection techniques ahead of any real-world execution.

4. Ships defenses for attacks that do not yet exist: By September, Innerworks will be protecting clients against attack categories that no threat intelligence feed can currently name.

Piloting is already underway against the live Bot Virus challenge, which is why the current subnet focus sits on bot detection rather than the geolocation or fingerprinting lanes.

Why This Cannot Be Built Anywhere Else

Cybersecurity today runs as a black-box duel where both sides operate blind.

1. Attackers guess at Cloudflare-tier defense internals: They iterate against assumptions about how those systems work.

2. Defenders guess at attacker behavior from collection code alone: They interpret partial signals and hope the interpretation holds.

3. Both sides waste effort on wrong assumptions: Neither side sees what the other is doing until damage happens.

RedTeam solved half the problem by receiving exact attack scripts from miners. Those scripts flow automatically through code ingestion, technique identification, and defense deployment.

The Immune System closes the gap by exposing defense internals to attacks. That access lets Innerworks construct threats the market may miss for years.

The Commercial Base Behind the Launch

The September release lands on top of a customer base that gives the announcement immediate credibility.

1. Five paying customers in production, 125 million+ daily active users protected: 1inch is the only publicly disclosed name, with everyone else hidden behind security-industry confidentiality.

2. Five enterprise pilots underway with a combined $22B+ in valuation: Nearly a billion accounts reached, over a billion weekly transactions handled between them.

3. Detection curve outperforming every named competitor: RedTeam’s true-positive rate climbed steadily while incumbents fell from 60% detection to roughly 15% over the same period.

4. Device fingerprinting (75% fewer collisions than the top competitor): Two similar MacBooks getting merged into a single identity is the failure mode that ruins fraud models, and RedTeam’s fingerprinting keeps them separate.

5. Geolocation (100% accuracy against a top compliance provider’s 40% under VPN traffic): 60% of users were being misattributed to the wrong country because their VPN exit node happened to sit somewhere else.

The company is fully committed to the subnet through Conviction, and generates no sell pressure on subnet ownership because revenue from clients funds buybacks directly.

September Is the Deadline

The Immune System is targeting September, with today’s bot virus challenge already testing production components. Its release creates a cybersecurity category incumbents cannot easily enter without exposing their defenses publicly.

Innerworks combines protected defense internals with a global adversarial contributor network through Bittensor. The six month sprint therefore turns a theoretical advantage into a commercial cybersecurity product by September.