‘Too Dangerous’: Anthropic Mythos Sparks Crypto Hack Fears Amid Coinbase Talks

Anthropic has launched a new artificial intelligence model that is causing ripples in the crypto industry. Security experts are concerned that the new Anthropic Mythos model may become the fastest way to launch a cyberattack on exchanges. It could, in turn, lead to rise in crypto hacks. However, Coinbase has indicated interest in adopting this AI Model despite these fears.

Why The Anthropic Mythos AI Model Could Trigger Crypto Hacks

Anthropic’s “Claude Mythos Preview” model has been described by developers as “too dangerous” for public release. The reasons include its ability to autonomously detect and exploit software vulnerabilities. Crypto analyst Ali Martinez said that it boasts an “autonomous capacity to discover and chain thousands of zero-day vulnerabilities.”

The apprehensions come at the time when the crypto industry is yet to overcome previous losses. According to data from Chainalysis 2026 Crypto Crime Report, over $3.4 billion of the crypto assets were stolen in 2025.

For further context, last year, the theft volumes hit historic highs in only 142 days. In addition, exploits of high profile systems such as a $1.5 billion hack on Bybit have brought increased focus on platform security. Also in April 2026, Drift Protocol faced a hack with Circle sued over not freezing $230 million USDC.

Moreover, the capabilities of Anthropic Mythos are more than traditional vulnerability scanning. It is able to detect bugs, auto-generate exploit code and run chains of attacks. During a single internal test, the model allegedly found a decades old bug in OpenBSD and used consecutive vulnerabilities to circumvent browser lockouts. Such tasks often demand elite security teams.

Further, the magnitude of risk has gained spotlight due to recent incidents. An insidious application that impersonated Ledger Live bypassed the review process of Apple and stole $9.5 million of users. In addition, Rhea Finance lost $7.6 million in a scam of oracle manipulation. In a different incident, the Hyperbridge attack cost the company $2.5 million due to a replay vulnerability.

According to market participants, centralized platforms might be the key targets after Anthropic Mythos launch. Crypto exchanges such as Coinbase, Robinhood, and Gemini process massive amounts of money and personal information, which could provide easy access to attackers.

“Any other system that deals with money in a real-time basis is going to be a place that we try to look for cyber security holes,” according to Cosimo Jiang of Pantera Capital. Analysts also raised concerns on reputation risks. They noted AI tools could even amplify phishing campaigns and synthetic identity fraud.

Coinbase & Binance Move To Adopt AI Technology

Meanwhile, big exchanges like Coinbase and Binance are considering defensive uses of the AI technology. Coinbase affirmed that it is negotiating with Anthropic and believes that there is a future of applying sophisticated AI to test systems.

“Mythos, and future models like it, will enable even deeper testing of software and systems at scale. This will accelerate digital threats as well as digital defense,” said Philip Martin, Chief Security Officer at Coinbase, per a CNBC report.

On the same note, Binance announced that it is working on AI to detect vulnerabilities more effectively throughout its infrastructure. “Near term, it will become a negative narrative for these kinds of companies,” said Jimmu Su, Chief Security Officer at Binance. He added, “But longer term, I would see them as one of the first batches that comes out and can protect against these AI agents.”