Author here. Spent the last few weeks chasing down the AI-attributed attack cases that made the rounds this year, including the Mexican government breach, the "vibe hacking" story, the Algerian amateur. Basically trying to work out whether hacking is impacted by broader AI adoption or whether the press was running ahead of the evidence.
On one side, Daniel Stenberg ran the gated Anthropic frontier model against curl on May 11. Five "confirmed" findings, one low-severity CVE after triage. His words: "the big hype around this model so far was primarily marketing." Stenberg is not a guy who hedges, and curl is not a toy codebase.
On the other side, there's SCONE — Anthropic's own December 2025 benchmark. Agents exploited 19 of 34 post-cutoff smart contracts, 55.8% success, $4.6M in simulated funds at an average API cost of $1.22 per contract. The comparable number 12 months earlier was about 2%.
Looks like agents are getting genuinely good at narrow, well-scoped vulnerability classes (Solidity, post-cutoff, bounded targets) and still bad at messy real-world codebases. But that's a guess and I'd rather hear pushback.
Happy to get into methodology, the spots where Chainalysis, Immunefi, and Web3IsGoingJustGreat don't line up, or specific cases. 28 references at the end of the piece.
> On the other side, there's SCONE — Anthropic's own December 2025 benchmark. Agents exploited 19 of 34 post-cutoff smart contracts, 55.8% success, $4.6M in simulated funds at an average API cost of $1.22 per contract. The comparable number 12 months earlier was about 2%.
Anthropic has a vested interest in making their LLMs look advanced, powerful and dangerous. This is the company that is explicitly pro-regulation, who has donated $20M to a PAC for pro-regulation candidates, and whose own competitors accuse of being pro-regulatory capture. We should take their benchmarks and their "Mythos is too dangerous for you mere mortals" statements with a big ass grain of salt, because it plays directly into that regulation angle they're playing. Anthropic wants frontier model development locked up, with only a few select stewards of humanity holding the keys.
You wrote the blog and this comment with Claude Opus.
I'm sure you meant well and only used it for editing, etc. etc., and I agree AI is good.
In any case, I can't trust AI on AI, especially with such a stark headline from someone outside Anthropic. (how do you know it was a solo user with Claude?)
This is either breaking news that you for some reason delegated to an overly verbose post written by AI, or, its an almost-true-but-not-quite clickbait title, and I don't have the domain chops to know. Impossible spot to be in as a reader.
References
[1] SecurityWeek, "Hackers Weaponize Claude Code in Mexican Government Cyberattack," Feb. 2026. [Online]. Available: https://www.securityweek.com/hackers-weaponize-claude-code-i...
[2] Anthropic, "Threat Intelligence Report: August 2025," Anthropic, Aug. 27, 2025. [Online]. Available: https://www-cdn.anthropic.com/b2a76c6f6992465c09a6f2fce282f6...
[3] D. Stenberg, "Mythos finds a curl vulnerability," daniel.haxx.se, May 11, 2026. [Online]. Available: https://daniel.haxx.se/blog/2026/05/11/mythos-finds-a-curl-v...
[4] Trail of Bits and OpenZeppelin, "Arbitrum Research and Development Collective (ARDC) procurement-grade pricing benchmarks," 2024. Approximately $25,000 per engineer-week for senior smart-contract auditing.
[5] W. Xiao, C. Killian, H. Sleight, A. Chan, N. Carlini, and A. Peng, "AI agents find $4.6M in blockchain smart contract exploits," Anthropic Red Team / MATS / Anthropic Fellows program, Dec. 1, 2025. [Online]. Available: https://red.anthropic.com/2025/smart-contracts/
[6] P. Paganini, "Claude code abused to steal 150GB in cyberattack on Mexican agencies," SecurityAffairs, Feb. 2026. [Online]. Available: https://securityaffairs.com/188696/ai/claude-code-abused-to-...
[7] Immunefi, "2026 State of Onchain Security," Immunefi, Jan. 2026. 425 publicly disclosed exploits 2021-2025 totaling $11.9 billion; cumulative whitehat payouts exceed $110 million across 330+ projects and 45,000+ researchers.
[8] Chainalysis, "2026 Crypto Crime Report," Chainalysis, Feb. 2026. 2025 stolen funds totaled $3.4 billion; cumulative DPRK take all-time, $6.75 billion.
[9] M. White, "Web3 Is Going Just Great," web3isgoinggreat.com. (Cumulative loss tracker, broader scope including exchange and protocol collapses.) [Online]. Available: https://web3isgoinggreat.com
[10] Z. Wang, X. Chen, Y. Chen, et al., "Characterizing Ethereum Upgradable Smart Contracts and Their Security Implications," arXiv:2403.01290, Mar. 2024. (Measurement study covers 60,251,064 Ethereum smart contracts.) [Online]. Available: https://arxiv.org/abs/2403.01290
[11] Flipside Crypto, "EVM Layer-2 deployment statistics," Flipside Crypto, 2024. More than 637 million EVM contracts across 7 L2 chains; Optimism alone hosted approximately 70% in 2024 YTD.
[12] Etherscan, "Daily Verified Contracts Chart," etherscan.io. All-time peak of 602 verified Solidity contracts deployed in a single day in 2023. [Online]. Available: https://etherscan.io/chart/verified-contracts
[13] Google Project Zero and Google DeepMind, "From Naptime to Big Sleep: Using Large Language Models To Catch Vulnerabilities In Real-World Code," Google Project Zero, Oct. 2024. [Online]. Available: https://projectzero.google/2024/10/from-naptime-to-big-sleep...
[14] N. Perry, M. Srivastava, D. Kumar, and D. Boneh, "Do Users Write More Insecure Code with AI Assistants?" in Proc. 2023 ACM SIGSAC Conference on Computer and Communications Security (CCS '23), Copenhagen, Denmark, Nov. 2023. 47 Stanford participants on codex-davinci-002. [Online]. Available: https://arxiv.org/abs/2211.03622
[15] United States v. Eisenberg, No. 23 Cr. 10 (S.D.N.Y. May 23, 2025), Opinion and Order on Rule 29 Motion for Acquittal (Subramanian, J.), 35 pp. [Online]. Available: https://nysd.uscourts.gov/sites/default/files/2025-05/23cr10...
[16] E. Calvano, G. Calzolari, V. Denicolò, and S. Pastorello, "Artificial Intelligence, Algorithmic Pricing, and Collusion," American Economic Review, vol. 110, no. 10, pp. 3267-3297, Oct. 2020. [Online]. Available: v
[17] S. Fish, Y. A. Gonczarowski, and R. I. Shorrer, "Algorithmic Collusion by Large Language Models," arXiv:2404.00806, Apr. 2024. [Online]. Available: https://arxiv.org/abs/2404.00806
[18] CoinDesk, "Attacker Drains $182M From Beanstalk Stablecoin Protocol," Apr. 17, 2022. See also PeckShield and Omniscia post-mortems documenting the flash-loan governance attack and emergencyCommit exploitation of BIP-18. [Online]. Available: https://www.coindesk.com/tech/2022/04/17/attacker-drains-182...
[19] The Block, "$24 million Compound Finance proposal passed by whale over DAO objections," Jul. 29, 2024. Proposal 289 vote: 682,191 in favor, 633,636 against. [Online]. Available: https://www.theblock.co/post/307943
[20] DARPA, "AI Cyber Challenge marks pivotal inflection point for cyber defense," DARPA, Aug. 2025. Team Atlanta (Georgia Tech, KAIST, POSTECH, Samsung Research) won the $4 million top prize with the ATLANTIS cyber-reasoning system; 54 of 63 synthetic vulnerabilities discovered (86%) and 43 patched (68%) across 54 million lines of code. [Online]. Available: https://www.darpa.mil/news/2025/aixcc-results
[21] CETaS, "Claude Mythos: What Does Anthropic's New Model Mean for the Future of Cybersecurity?" Centre for Emerging Technology and Security, The Alan Turing Institute, Apr. 2026.
[22] Anthropic, "Responsible Scaling Policy v3.0," Anthropic, Feb. 2026.
[23] European Parliament and Council of the European Union, "Regulation (EU) 2024/1689 of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (AI Act)," Official Journal of the European Union, Jul. 12, 2024. Dual-use provisions in next implementation phase scheduled for August 2026.
[24] National Institute of Standards and Technology, "AI Risk Management Framework (AI RMF 1.0)," NIST AI 100-1, Jan. 2023. [Online]. Available: https://www.nist.gov/itl/ai-risk-management-framework
[25] AI Safety Institute (UK), "The Last Ones: 32-Step Corporate-Network Attack Simulation," AI Safety Institute, Apr. 2026.
[26] V. Buterin, "The Promise and Challenges of Crypto + AI Applications," vitalik.eth.limo, Jan. 30, 2024. [Online]. Available: https://vitalik.eth.limo/general/2024/01/30/cryptoai.html
[27] Lido DAO, "Dual Governance — Lido Improvement Proposal LIP-28," Lido Finance. Activated on Ethereum mainnet, Jun. 30, 2025. 1% TVL "first seal" threshold and 10% TVL "rage-quit" threshold. Built with audits by Certora, OpenZeppelin, Statemind, and Runtime Verification; agent-based simulations by Collectif Labs; game-theoretic models by 20squares. [Online]. Available: https://github.com/lidofinance/lido-improvement-proposals/bl...
[28] Anthropic, "Disrupting the First Reported AI-Orchestrated Cyber Espionage Campaign (GTG-1002)," Anthropic, Nov. 13, 2025. Approximately 30 targets across technology, finance, chemicals, and government sectors. [Online]. Available: https://www.anthropic.com/news/disrupting-AI-espionage
Looked at [1] and [6] and yeah, it wasn't a solo user with just Claude Code. And the sources are garbage lol, both are rewrites of a startup called Gambit's press release. I'm surprised Claude wasn't more careful, to be honest, the articles stop far shy of "solo user with Claude Code" and provide more context that obviates it.
Thanks for adding to your blog retroactively I suppose (version history would almost be nice).
Feel my time was wasted today and will share “Slop is something that takes more human effort to consume than it took to produce.”
—
Interesting, wonder if it was solo or team hoping to appear solo
On one side, Daniel Stenberg ran the gated Anthropic frontier model against curl on May 11. Five "confirmed" findings, one low-severity CVE after triage. His words: "the big hype around this model so far was primarily marketing." Stenberg is not a guy who hedges, and curl is not a toy codebase.
On the other side, there's SCONE — Anthropic's own December 2025 benchmark. Agents exploited 19 of 34 post-cutoff smart contracts, 55.8% success, $4.6M in simulated funds at an average API cost of $1.22 per contract. The comparable number 12 months earlier was about 2%.
Looks like agents are getting genuinely good at narrow, well-scoped vulnerability classes (Solidity, post-cutoff, bounded targets) and still bad at messy real-world codebases. But that's a guess and I'd rather hear pushback. Happy to get into methodology, the spots where Chainalysis, Immunefi, and Web3IsGoingJustGreat don't line up, or specific cases. 28 references at the end of the piece.