Cardano Founder Highlights Key Solution to Growing AI Threats After Liquid Network’s $320M Bitcoin Breach

Cardano founder Charles Hoskinson has reacted to the reported theft of 4,000 Bitcoin from the Liquid Network’s federation wallet.

Liquid confirmed the incident in an X post yesterday, stating that hackers siphoned approximately 4,000 Bitcoin worth $320 million from its federation wallet. According to the network, the attackers moved the funds through the SideSwap Peg-out Authorization Key (PAK) service.

However, Liquid and SideSwap both indicated that the PAK itself was not compromised. Instead, SideSwap said the L-BTC involved in the transaction originated from a vulnerability in Elements, the Bitcoin-derived software that powers the Liquid sidechain.

Following the incident, Liquid disabled its bridge nodes and paused network activity while federation members investigated the breach. Meanwhile, exchanges received notifications and began pausing, or preparing to pause, L-BTC deposits and withdrawals.

The transaction also contained an OP_RETURN message, in which the attackers, who claimed to be white hats, requested on-chain contact.

Hoskinson Warns of AI-Driven Cybersecurity Risks

Reacting to the incident, Hoskinson used the breach to highlight what he sees as a growing cybersecurity threat from artificial intelligence.

He argued that formal methods may provide one of the strongest defenses against increasingly capable AI-driven attacks. As AI systems become better at analyzing source code, identifying vulnerabilities and automating sophisticated attacks, blockchain developers face an increasingly narrow window between the discovery of a software weakness and its exploitation.

AI has already begun changing the cryptocurrency security landscape by helping attackers identify potential targets, analyze complex code, and automate phishing and social-engineering campaigns. The reported Coldcard hack, which involved the theft of $130 million worth of Bitcoin, has further fueled concerns about how advanced AI tools could make difficult vulnerabilities easier to identify and exploit.

Against this backdrop, Hoskinson argues that conventional testing alone may no longer provide sufficient protection. Instead, he points to formal methods, which use mathematical specifications and proofs to establish whether software behaves according to its intended design.

Cardano’s Emphasis on Formal Verification

Hoskinson’s warning also aligns with Cardano’s long-standing emphasis on mathematically rigorous blockchain development.

Cardano’s development philosophy incorporates academic research, peer review, and formal verification as important components of its security approach. Its Ouroboros consensus protocol, for example, emerged from formal academic research, while the broader ecosystem has emphasized proving critical properties of protocols and software rather than relying exclusively on experimental testing.

This approach supports Hoskinson’s argument because formal verification seeks to mathematically demonstrate that specified properties hold. As a result, developers can potentially identify entire classes of vulnerabilities before deploying code to production.

Cardano also relies on functional programming technologies such as Haskell and Plutus. Their strong type systems and mathematical foundations can help developers detect certain categories of programming errors earlier in the development process.

Notably, Cardano’s approach aims to prevent vulnerabilities before they reach production rather than relying solely on testing and conventional security practices.