# Security Roundup: Mid-Six-Figure DeFi Drains Continue as Researchers Test AI Against On-Chain Exploits

Fresh proof-of-concept reconstructions document a $222,560 stablecoin drain and a smaller lending exploit, while a new benchmark tests whether artificial intelligence (AI) agents can secure contracts.

- Published: 2026-06-26T10:31:33.042Z
- Canonical: https://polylog.news/crypto/2026-06-26/security-roundup-mid-six-figure-defi-drains-continue-as-rese
- Publisher: Polylog (Crypto desk)
- Section: crypto
- Sources: [DeFiHackLabs](https://github.com/SunWeb3Sec/DeFiHackLabs/commit/476c1ec32a4542bb9d3acc567a7c30b94036ecf7), [DeFiHackLabs](https://github.com/SunWeb3Sec/DeFiHackLabs/commit/e7a2c7837a5a31d1c39a08141e0ebba0a6164e14), [arXiv](https://arxiv.org/abs/2606.26216)

The steady series of smaller decentralized finance (DeFi) exploits continued this week. Security researchers at DeFiHackLabs published a reconstruction of the DLMC incident, a drain of about 222,560 USDT dated June 24, alongside a proof-of-…

This story is for subscribers. Read it in full at https://polylog.news/crypto/2026-06-26/security-roundup-mid-six-figure-defi-drains-continue-as-rese (subscription information: https://polylog.news/pricing).