Anthropic warns AI worms could spread across consumer devices

Traders Union
All News
Crypto
Anthropic warns of AI worms

Highlights

Anthropic researchers warn that AI-powered worms may soon self-replicate and spread autonomously across IoT devices with weak security, challenging containment efforts.
One-third of the Bitcoin supply (6.26 million BTC) is viewed as vulnerable to AI and quantum computer attacks, with 4.33 million exposed due to address reuse, per Glassnode.
Ethereum researcher Justin Drake cautions that rapid AI advancement could soon break the ECDSA cryptographic standard protecting crypto wallets, urging users to migrate funds.

Research flags self-replicating device threat

As reported by Cointelegraph, citing Anthropic, researchers William Zhou and Daniel Freeman from the company’s Frontier Red Team warn that rogue AI worms may begin spreading across user devices without human intervention, in a pattern similar to computer viruses.

The report says such systems may target Android TVs, smart appliances, cameras and other internet-connected devices that combine modest computing power with limited security protections. The authors add that the threat could prove difficult to contain because a self-replicating network of compromised devices would have no central switch to turn off.

Anthropic’s researchers say small AI models may be months away from becoming capable enough to operate in this way on commercial IoT hardware, raising the prospect of a decentralized cyber risk tied to rapid advances in model capability.

Crypto security concerns widen

The warning comes as debate intensifies over whether stronger AI systems could also undermine cryptographic protections used in digital assets. The article says roughly one-third of the Bitcoin supply, or 6.26 million BTC, currently sits in addresses seen as vulnerable to future attacks from AI models or quantum computers.

Glassnode co-founder Rafael Schultze-Kraft wrote in a Thursday post on X that 4.33 million of those coins are exposed because of address reuse, meaning the risk could be reduced by moving Bitcoin to fresh wallets. A day earlier, Ethereum researcher Justin Drake warned that rapid AI progress may break the elliptic curve digital signature algorithm, or ECDSA, that secures crypto wallets sooner than previously expected, and urged users to migrate funds.

Not all industry figures agree with that assessment. Dragonfly managing partner Haseeb Qureshi called the warning cryptographic doomerism and argued that moving tokens to new addresses does not offer a real solution to the broader threat.

Our earlier article on CTOS Digital Bhd’s cybersecurity incident detailed how the disruption began translating into measurable financial damage, with the company estimating a hit to FY2026 profit and revenue if services stayed affected into October. We also noted the market fallout, including a sharp share-price decline, alongside CTOS’s recovery steps such as added security measures and an independent review.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.