Zcash

<h2 id="header-0">Key Facts</h2>
<ul>
 <li>Bill Gates said AI has crossed key bio and cyber danger thresholds, and that the technology "makes nuclear weapons look like nothing."</li>
 <li>Gates said relying on the AI industry to regulate itself is "just insane" in an interview published the same day tech CEOs signed a voluntary safety accord at the White House.</li>
 <li>A security engineer using Anthropic's Opus 4.8 found a Zcash flaw on May 29 that had been live for four years, writing a working exploit that minted counterfeit ZEC in testing.</li>
</ul>
Bill Gates said artificial intelligence has crossed the thresholds that make it more dangerous than any weapon humanity has built, and that almost no one responded when it happened.
"This makes nuclear weapons look like nothing," Gates <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.youtube.com/watch?v=A_156w0aYtU">told</a> the Ezra Klein Show in an interview published Tuesday. "We crossed those thresholds, and there was complete silence."
The Microsoft co-founder said the capability to "empower a bioterrorist to kill hundreds of millions" exists today, as does to carry out "a cyberattack that scrambles all the bank accounts, shuts down the electric grid."
<h2 id="header-1">What Convinced Gates AI Crossed the Cyber Threshold</h2>
Gates said he sees the evidence in AI’s code abilities. He said recent models began finding flaws "in code that humans have looked over for over 20 years and said, boy, we see there's no problem here."
"And it, in a few minutes says, no, no, I can inject this over here and this over here," he said. "At a level of complexity where when you see it, you go, wait, oh yeah, you're right."
"It used to be only nation states had enough resources and capability to do these things," he said. "Now that power has been passed into the hands of a small group just using the latest AI tools."<tweet-embed url="https://x.com/BillGates/status/2105084993827704946"></tweet-embed>
<h2 id="header-2">Has AI Been Used in a Crypto Hack Yet?</h2>
AI’s capability to disrupt crypto is already documented. Security engineer Taylor Hornby used Anthropic's Opus 4.8, released the day before, in an <a target="_blank" rel="noopener noreferrer nofollow" href="https://coinmarketcap.com/academy/glossary/audit">audit</a> of Zcash's (<a target="_blank" rel="noopener noreferrer nofollow" href="https://coinmarketcap.com/currencies/zcash/">ZEC</a>) Orchard circuit on May 29.
He found a flaw live since Orchard launched in May 2022 and wrote an exploit that minted undetectable counterfeit ZEC in testing, <a target="_blank" rel="noopener noreferrer nofollow" href="https://shieldedlabs.net/the-orchard-counterfeiting-vulnerability/">Shielded Labs</a> said. An emergency fix followed within days.
The flaw "had evaded years of scrutiny by many of the world's best cryptographers," Shielded Labs said. Anthropic's December 2025 <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.anthropic.com/research/smart-contracts">study</a> found AI agents identified $4.6 million of exploitable <a target="_blank" rel="noopener noreferrer nofollow" href="https://coinmarketcap.com/academy/glossary/smart-contract">smart contract</a> vulnerabilities.
Whether attackers have used AI in real-world exploits is not yet established, but it is high suspected. CertiK put first-half losses at $1.32 billion and <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.certik.com/skynet-report/certik-hack3d-h1-2026-report">said</a> attackers revisiting dormant codebases were "likely assisted by automated tooling."
Blockstream, <a target="_blank" rel="noopener noreferrer nofollow" href="https://coinmarketcap.com/academy/article/blockstream-refuses-ransom-bitcoin-stolen-liquid-network">refusing to pay a ransom</a> over the Sept. 6 Liquid Network theft, said "the world is a different place with the advancements of AI," hinting that AI could have been used in the exploit.
Asked whether corporate incentives were enough to keep humanity safe from AI, Gates said: "I almost can't believe you're asking that.”&nbsp;
<blockquote>
 “This is the most dangerous thing that humans have ever gone near."
</blockquote>
There is "no supervisory layer today," he added, calling self-regulation "just insane."
Read More: <a target="_blank" rel="noopener noreferrer nofollow" href="https://coinmarketcap.com/academy/article/ai-agents-crypto-how-to-get-started-2026">AI Agents in Crypto: How to Get Started in 2026</a>

Bill Gates Says AI Makes Nuclear Weapons 'Look Like Nothing'

Table of Contents

Other articles published on Oct 1, 2026

Trump Meme Coin Plans Third Gala Dinner for Top 185 Holders

Hyperliquid Policy Group Urges EU To Classify Perpetuals Under MiFID

Citi Lifts Bitcoin Target to $113K, Ethereum to $3,028

Join our free newsletter for daily crypto updates!