Crypto Network Crowdsources Bot Detection Through Competitive 'Ethical Hacking'

A cybersecurity project called RedTeam is using a decentralised computing network, Bittensor, to crowdsource bot detection improvements. Operated by security firm Innerworks, the project sets miners against each other in ongoing security challenges, with participants submitting code-based solutions that are tested in controlled environments and rewarded with tokens based on performance.

Over 11 months, this competitive approach reportedly pushed bot detection accuracy from 74.3% to 99.5%, while rival systems in the same tests saw their average performance drop to just 14.6%. The project has run 14 distinct challenges since late 2024, retiring nine as detection capability matured. Innerworks says its technology is now integrated with a decentralised finance protocol and a messaging app with over 100 million daily users, with five more large enterprise applications currently being tested.

While this is a specialised, large-scale industry development rather than a direct threat to small businesses, it signals that bot detection and fraud prevention technology is advancing rapidly, meaning the automated bots and scrapers targeting websites and login pages are also likely to keep evolving in response.

Key Takeaway: Small businesses should keep website and account security tools such as bot protection and multi-factor authentication up to date, as detection technology and the automated threats it targets are both evolving quickly.