Multiple security incidents have hit the crypto and AI sectors recently: MultiversX paused its mainnet after an atomicity vulnerability attack; Fetch.ai and NuNet lost $2 million in a same-source attack; Polymarket is mired in card fraud; Zhipu ZCode is accused of leaking its full source code.
Security incidents erupted across the crypto and AI sectors this week, affecting public blockchains, AI concept tokens, prediction markets, and AI coding tools, with clear signs of risk spillover.
MultiversX's mainnet was hit by an atomicity vulnerability attack, prompting the team to urgently suspend network operations and prepare fixes. As a major public blockchain, the attack has broad implications and directly threatens user asset security, and has been listed as a top-priority security event.
Following closely, leading AI concept project Fetch.ai and NuNet confirmed they were attacked by the same hacker, with total losses reaching $2 million. The attack put pressure on both FET and NTX token prices, with NTX plunging 65%. The incident has had a notable impact on AI-track crypto assets and is highly time-sensitive.
On the payment and compliance front, Polymarket, the largest prediction market platform in the U.S., has been exposed as deeply embroiled in card fraud. At its peak, over 80% of deposits on the platform were deemed chargebacks, far exceeding industry standards. The incident involves payment security vulnerabilities and compliance issues, materially affecting the platform's reputation and regulatory discussions in the industry.
The AI field also raised red flags. Zhipu ZCode was independently audited by a partner, which claimed that ZCode had uploaded complete enterprise source code and sensitive credentials in bulk, a serious breach. The impact has spread from the AI industry to the crypto and developer communities, sparking widespread discussion on agent data security audit mechanisms and drawing significant attention.