A group of 20 to 25 volunteers has formed a special force called the Bitcoin Red Team to hunt for security vulnerabilities using artificial intelligence. This emergency force consists of developers - many operating under pseudonyms - racing to find weaknesses in the crypto ecosystem before hackers do.
The initiative was triggered by a series of two significant events. The hack of the air-gapped Coldcard wallet served as a stark warning to developers. This situation was exacerbated by the release of China’s Kimi K3 AI model, an artificial intelligence that has reportedly caused widespread disruption in the cybersecurity world.
The team does not dissect Bitcoin’s core protocol code. Instead, their attention is fully focused on the application layer, storage wallets, complementary services, and surrounding supporting software. It is in these outer layers that security vulnerabilities are typically found.
Scanning Open-Source Code
The ranks of the Bitcoin Red Team include prominent developers such as Ben Carmen, Daniela Brozzoni, James O’Beirne, and Bruno Garcia, a board member of the Vinteum R&D Center. They operate on two fronts: servicing scan requests from various projects, while proactively scanning behind the scenes.
“We have scanned almost the entire significant open-source ecosystem,” said Calle, a developer and maintainer of the Cashu project who joined the team.
This hunt for vulnerabilities is costly. The team has spent around $20,000 so far to fund various security services. However, all operational funding has been secured, so additional donations from the public are not required.
AI Lowers the Barrier to Hacking
Artificial intelligence is changing the game because it lowers the technical barrier to exploitation to its lowest point. Someone can now execute a full hack using nothing but prompts to an AI engine, without needing any foundational skills.
“Simple exploits can now be completed end-to-end by someone who wouldn’t know how to do it without AI,” Calle explained.
He emphasized that the concept of hiding vulnerabilities through code complexity is dead. “There are no secrets in software anymore - there is no information asymmetry that was once used for security through obscurity. That is over,” he added.
Given its appeal, Bitcoin is high on the priority list for hackers. As an attacker, the first thing you want to target is internet money.
The Irony of Using Chinese Models
To run scans and simulated attacks, the team needs AI without limits. This is where they face obstacles with US-made models. The guardrails on American AI often block their activities because the engine flags them as exploits, even though they are researching security vulnerabilities.
“Although US models are generally smarter, their guardrails limit their usefulness in security,” Calle lamented.
As a workaround, they frequently rely on AI models built in China. These Asian models do not impose strict restrictions, allowing researchers the freedom to replicate hacker maneuvers without system interference.
As Calle emphasized, it is now only a matter of time. The Red Team exists because they must stay a step ahead of attackers as quickly as possible. This volunteer force carries the heavy burden of being the last line of defense separating automated hackers from user assets worldwide. Reported by Decrypt.
Disclaimer: This article is for informational and educational purposes only, not financial advice. Cryptocurrency assets are highly volatile and carry significant risk. Always do your own research (DYOR) and never invest more than you can afford to lose.
