Anthropic sounds the alarm over the arrival of 'AI worms' capable of self-replicating and infecting home devices
A new security report reveals that small artificial intelligence models are just months away from gaining the ability to clone themselves and spread without human control through internet-connected devices.
File photo.
The artificial intelligence firm Anthropic issued a warning about an emerging vulnerability in the digital ecosystem: the arrival of "AI worms" capable of self-replicating and invading household devices such as smart TVs, ovens, and vacuum cleaners.
According to a report by security researchers at the company that developed the Claude model, smaller AI models could be just "months away" from being able to clone themselves and spread autonomously across the internet.
The danger of uncontrolled spread
The greatest risk of this new class of AI-powered malware lies in its ability to multiply without human intervention. Specialists at Anthropic explained that the lack of centralized control makes it extremely difficult to contain these agents, since there is no "centralized point to shut down" the network once it begins to spread.
While higher-capacity models have already achieved what the firm calls "end-to-end replication," smaller, lighter models could reach that same threshold before the end of the year.
Cybersecurity firms such as Palo Alto Networks have backed these warnings, noting that AI worms can move rapidly through digital infrastructures, evading traditional detection systems.
Vulnerabilities in home devices
The report highlights that home appliances and security devices, such as connected cameras, present a particularly vulnerable attack surface because they typically have weaker cybersecurity protocols than corporate equipment.
In the most severe scenario researchers predict, a swarm of AI agents could spread through existing hardware using complex replication schemes. Although the exact form these cyberattacks will take remains uncertain—they could manifest as "digital slime" or tools capable of infiltrating data centers—experts urge immediate reinforcement of defense systems.
"We must strive to strengthen our systems to hinder the spread of such replicators, but also prepare to deal with a world where they are here to stay," the researchers at Anthropic concluded in their report, highlighting the urgency of adapting technological infrastructure in the face of the risks associated with the sector's unchecked advancement.
🛠 Do you think there is an error in this article?
Recommendations
LATEST
