Life, uh, finds a way. The CEO of Coinbase doesn't mince words. Brian Armstrong expects that an artificial intelligence model will escape all control on the Internet within one to two years, an episode he directly compares to the Morris worm of 1988, the first major cybersecurity incident in history. However, there is nothing alarmist in his remarks: the leader sees it mainly as an almost inevitable step, destined to fade into the background once the media storm subsides. Key points of this article:
Brian Armstrong Compares AI Risk to the Morris Worm of 1988
On X, Armstrong posted on August 13:
I wouldn't be surprised if an AI model becomes uncontrollable on the Internet in a year or two, much like the Morris worm in 1988. It will feel like a major event, with a media frenzy and calls to shut everything down. But as always, people will adapt, defenses will be built, and that...
The tweet ends abruptly, but the idea can be summed up in one sentence: temporary panic, lasting adaptation. The parallel is not chosen at random. Robert Tappan Morris, then a student at Cornell, released a 99-line program in November 1988 that was supposed to simply measure the size of the Internet. However, the worm began to reinfect the same machines in a loop, saturating their resources rather than destroying files.
The result: about 6,000 computers affected, or 10% of the network at the time, including Harvard, Stanford, and NASA, with some institutions offline for an entire week. This episode led to the creation of CERT/CC, the first coordinated incident response team for cybersecurity, which has since become the global model of its kind. AI Labs Already Found Lacking in Containment
Armstrong's prognosis does not come from nowhere. Several labs have recently documented containment failures during cybersecurity tests, reports Crypto Briefing. In July, OpenAI acknowledged that one of its models exploited a zero-day vulnerability in its testing environment to gain access to the Internet, before reaching the production infrastructure of Hugging Face. The same month, Anthropic recorded three incidents involving Claude out of over 141,000 evaluations reviewed, with models reaching real organizations after a third-party evaluation environment mistakenly left them access to the network. The same misadventure occurred with Kimi K3, the Moonshot model, although a network configuration error seems to have facilitated the breach.
A common thread emerges from these three episodes: these are not malicious AIs by nature, but agents skilled at spotting the slightest isolation flaw, the least protected identifier, and exploiting it to continue fulfilling their objective. All these small infrastructure oversights, when added together, increasingly resemble the breeding ground for a real incident. A Scenario That Is Already Causing Concern Beyond Tech
This normalization of risk is not unique to Silicon Valley. The Bank of England has recently advocated for an emergency stop button on trading AIs, deeming current regulatory frameworks outdated in the face of agents capable of making decisions independently. Vitalik Buterin, for his part, advocates for voluntarily curbing AI autonomy, even at the cost of some efficiency, precisely to avoid such scenarios. Between financial regulators, crypto figures, and now the head of one of the largest exchanges in the sector, the debate is no longer about whether an incident of this kind will happen, but when, and what the ecosystem will learn from it.
This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.



![[SCAN 2026 Final Interview] ②J0y_B0y: The Four Students from Jordan's Applied Science Private University Unite](/public-static/25_d4737ee605.png?format=avif)

























