BREAKING: AI Agents Can Now Execute Real Smart Contract Exploits—Anthropic Sounds Alarm
AI crosses the Rubicon—autonomous agents now actively exploit DeFi vulnerabilities.
Smart contracts meet smarter hackers
Generative AI isn't just writing poetry anymore. Anthropic's latest research reveals LLMs can now:
- Identify unpatched Solidity flaws
- Craft precision payloads
- Drain liquidity pools before you can say 'rug pull'
The ultimate crypto trading bot—except it trades your assets to itself. Hedge funds are reportedly licensing the tech while VCs scramble to invest in 'blockchain security AI' startups (too late, as usual).
One silver lining? Finally, a legitimate use case for those overpriced smart contract audits.
Anthropic applied AI models on real hacks and uncovered exploits worth $550M
They created SCONE-bench with 405 exploited contracts from 2020-2025, and 10 models crushed 51% of them
Is AI the new king of security audits? pic.twitter.com/spZWOst3hO
— CryptoLabs (@cryptolabsio) December 2, 2025
The study used Anthropic’s newly developed benchmark, analyzing. Leading AI models, includingand, were tested across a series of simulations based on data after March 2025.
According to Anthropic, AI agents successfully conducted attacks totaling, demonstrating that autonomous AI-driven exploits are no longer hypothetical.
AI Discovers Zero-Day Vulnerabilities and Generates Profit
On October 3, researchers evaluatedwith, many operating on.
Bothandidentified. These flaws were successfully exploited, allowing the AI to extract cryptocurrencies worth approximately.
The cost to operate GPT-5’s API for the attack was, meaning the exploit produced a.
Researchers emphasized that this marks a tipping point:.
The study focused not only on technical success, but also on the, underscoring the urgent need for stronger risk controls in AI-crypto infrastructure.
AI Attack Capabilities Doubling Every 1.3 Months
Anthropic found that the profitability of leading AI models in simulated attacks has been, driven by improvements in:
- autonomous tool use
- error recovery
- sequential reasoning
- complex exploit execution
While researchers noted that this explosive growth may eventually slow, they warned that the current pace represents a.
Anthropic concluded thatand that blockchain ecosystems urgently needto match accelerating offensive capabilities.
Interest is also growing in, including specialized crypto-security AI tools designed to monitor and block attacks in real time.
The post AI Agents Now Capable of Real Smart Contract Hacks, Anthropic Warns appeared first on icobench.com.
