Revenue Search: Inside Bittensor Subnet Session with Red Team: Beating Bots with Bittensor (SN61)
Jul 16, 2025
Oli and Oscar from Red Team dive into the rising threat of synthetic fraud driven by GenAI. They explain how their innovative decentralized R&D model leverages miners to simulate cyberattacks for better defense. Learn about the flaws in traditional bot detection and why CAPTCHAs are no longer effective. They discuss their B2B SaaS solution targeting high-volume payment platforms and explore the expansive market opportunities in crypto security. The podcast unveils their unique strategy and vision for the future of threat intelligence.
AI Snips
Chapters
Transcript
Episode notes
Three Core Signals For Transaction Security
- Red Team (Innerworks) builds signals to determine if a user is human, identify the physical device, and verify geolocation to protect financial transactions.
- These three signals are increasingly hard due to democratized GenAI tools that can simulate users and devices.
Adversarial R&D Powers Defense
- Innerworks uses an adversarial research approach: they attack detection systems to reverse engineer bypasses and improve defenses.
- That offensive R&D yielded the idea to decentralize attacks via BitTensor miners to expand coverage and diversity.
Accessibility Trumps Sophistication In Fraud
- The barrier to mount effective fraud is low: accessible LLMs plus prompt engineering let non-technical actors bypass many bot-detection products.
- The problem is accessibility and scale of attackers, not only cutting-edge sophistication.