[REVISION HISTORY]

AI cryptocurrency project exploits

Updated 1 time since CLSTR started tracking revisions of this situation.

What changed

2026-09-21 09:23 UTC → 2026-09-21 20:03 UTC · added ~~removed~~

A series of coordinated exploits targeting AI-focused cryptocurrency projects, including Fetch.ai, NuNet, and SingularityNET, resulted in significant financial losses and market volatility. Initial reports indicated that a single attacker targeted multiple platforms. In the Fetch.ai exploit, the attacker used compromised signing credentials to drain approximately 8.7 million FET tokens. This was followed by an attack on NuNet, where the perpetrator unauthorizedly minted 408.5 million NTX tokens, representing about 42% of the total supply. The most severe impact occurred on SingularityNET, where the illegal minting of AGIX and WMTX tokens caused the AGIX token to lose over 99% of its value within 24 hours, wiping approximately $93 million from its market capitalization. Subsequent analysis identified the breach as a compromise of a signing key linked to the Fetch.ai and SingularityNET bridge. Attackers exploited the TokenConversionManagerV3 contract, which connects Ethereum and Cardano. The breach appears to have occurred within the signing service or key custody rather than through a flaw in the contract code itself. The impact was exacerbated by design weaknesses, specifically a lack of per-conversion limits and a signed digest that did not bind the recipient address, allowing a valid signature to be used for any destination. The attacker also utilized a NuNet minter key that had been dormant since March 2023. Security firms PeckShield and Blockaid confirmed the single-attacker theory. While Fetch.ai stated its own contracts remained unaffected, the company paused AGIX-to-FET conversions as a precaution. The attacker is reported to hold tens of millions of dollars in various crypto assets following the incident.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.