Skip to main content

Hackers used autonomous AI agents to attack Taiwan. Is this the future of cyberwarfare?

(CNN) — Hackers deployed an AI system to carry out sophisticated cyberattacks on Taiwan, officials said, in what experts believe is the first known fully autonomous attack on government agencies.

Over four days in July, the AI agents were able to map 21 government systems, crack 85 government user accounts, and extract 2,500 personnel records, according to Dream, an Israeli AI firm that first discovered the attack.

Experts suspect the hackers are from China, but neither Taiwan, nor Dream would confirm the origin of the July attacks.

“The investigation found clear indications that the attacks originated overseas and involved a hybrid approach in which hackers combined conventional operations with AI agents such as OpenClaw,” Taiwan’s Ministry of Digital Affairs said in a Thursday statement.

The hackers’ system was built using open-source AI agents to automate and coordinate much of the intrusion, including reconnaissance, credential attacks and strategizing subsequent attack paths, Dream said. The Financial Times first reported the case on Wednesday.

The attack comes amid an increasing number of reports of advanced AI models engaging in unauthorized actions, raising fears of the proliferation of AI-enabled cyberattacks and prompting calls for more government action to regulate AI.

According to Kenny Huang, chairman of the Taiwan Network Information Center, this is believed to be the first disclosed case of a fully automated attack against a government.

Using AI to write code and search for vulnerabilities in cyberattacks has become common. But this incident went significantly further.

In this case, an autonomous system coordinating up to eight AI agents – carrying out intrusions and deciding what to do next without human intervention – effectively ran a hacking campaign rather than just merely assisting a human hacker.

“It spells out one thing loudly – the cost of running a competent attack has collapsed, but the cost of defending against one has not,” a blog post by Dream said.

Amir Becker, Dream’s chief business and strategy officer, said the way the AI system operated in this latest case was alarming.

“Like a human team, when an approach gets blocked, it researches new techniques in real time and adapts. It’s an attacker that strategizes, learns, and adjusts on its own,” he said.

The attack also extended to Taiwan’s nuclear safety agency and the government’s IT vendors as well as at least seven other energy sector companies.

The AI agents were able to quickly combine different hacking techniques and exploit vulnerabilities in secondary systems, allowing the attacks to be carried out faster, more cheaply and on a much larger scale, said the Taiwanese digital affairs ministry.

The use of simplified Chinese, used in China, found in internal documents linked to the hack suggests a high probability that the instigator, whether public or private, is linked to China, experts said.

CNN has reached out to China’s Ministry of Foreign Affairs, Taiwan Affairs Office, and Cyberspace Administration for comment.

Taiwan is no stranger to cyberattacks, particularly from neighboring China, which claims the East Asian democracy as its own and has vowed to annex it, by force if necessary. Last year, Taiwan was subject to an average of 2.6 million cyberattacks a day from China, a 6% increase from 2024, according to a government report.

As Beijing has stepped up its pressure campaign against Taipei in recent years, Taiwanese officials have said their democracy stands at the forefront of China’s “hybrid warfare,” facing disinformation campaigns, cyberattacks and near-daily military exercises around Taiwan.

Huang of the Taiwan Network Information Center said the July incident demonstrated that AI has become the main player in cybersecurity beyond simply assisting human hackers in partial automation tasks.

He said it raised the question of whether the world was ready with defensive strategies for AI, including the necessary legal frameworks, technical capabilities and policies.

“I believe there are still significant gaps. Every country, not just Taiwan, is still unprepared in this respect,” he said.

The-CNN-Wire
™ & © 2026 Cable News Network, Inc., a Warner Bros. Discovery Company. All rights reserved.

AI boom or bubble? Timing is everything

New York (CNN) — Artificial intelligence is a game-changing technology that’s already transforming how the world lives and works.Yet AI can also fuel a market meltdown, a recession – or both.Timing is everything. Money is pouring into AI so rapidly that it’s outpacing the ability to turn a profit. This unsustainable mismatch must come into balance, one way or another. The recent blow-up of a wildly successful AI-focused hedge fund demonstrates the risks and rewards of the AI boom.Those risks are magnified by how incredibly expensive it is to build the AI revolution, where companies are scrambling to buy cutting-edge chips and build data centers the size of dozens of football fields.“I absolutely believe the technology is transformative. But that doesn’t mean you won’t go through irrational exuberance at some point,” Max Gokhman, head of AI and digital asset solutions at investment firm Franklin Templeton, told CNN.Timing is critical in deciding whether the AI boom endures – or if it ends in tears like past asset bubbles.Will the gobs of money being spent to build out AI bring real returns before Wall Street’s patience runs out?$500 billion financing arrangementFor now, the heavy hitters in finance are still going headfirst into AI.Nvidia, the $5 trillion AI infrastructure superstar, just arranged $500 billion in financing from Apollo, BlackRock, Goldman Sachs and other Wall Street firms to bankroll customer orders for its cutting-edge chips.Nvidia CEO Jensen Huang even argued that AI compute – the hardware and software underpinning AI models – is transforming into an “investable class.”It’s a sign that many on Wall Street still believe in an AI-driven future. Investing in the technology has been complicated by its rapid developments; an emerging price war between open and closed AI models; and many supply chain bottlenecks – from access to chips and local opposition to data centers to the enormous power needs of the buildout.‘Will end badly’Nvidia’s blockbuster deal is a reminder of an innovative, and arguably dangerous, characteristic of the AI boom: circular financing.In circular financing, one company pays money to another (in the form of a loan, investment, lease or other financial support) in exchange for that second company buying the first’s products.It’s one of those things that works – until it doesn’t.During booms, these arrangements can create a virtuous circle. But they can also fuel speculative bubbles by creating the illusion of rapid growth — and those bubbles eventually burst.It brings back bad memories of the dotcom bubble, when some telecom equipment companies lent money to customers to buy their gear.“Circular financing will end badly,” Gokhman said. “You are living on not just borrowed time, but levered time.”But Gokhman said he’s still a believer in the AI boom and doesn’t think leverage has gotten to alarming levels, at least not yet.AI-focused hedge fund goes bustAs in the late 1990s, the AI boom has allowed some investors to make fortunes, even as others experience sudden disaster.Late last month, Situational Awareness, a little-known hedge fund focused on AI investments, imploded after hits highly leveraged bets blew up.Run by a former OpenAI employee, Situational Awareness was forced to sell most of its portfolio at a steep discount to Citadel. The AI hedge fund is still up big over the past two years – just not as much as it had been.Tellingly, Situational Awareness didn’t run into trouble because it was wrong directionally on its belief that AI is for real.The problem was that thesis got derailed by a temporary loss of momentum in AI stocks – and that derailment was magnified by leverage, which magnifies both gains and losses.AI’s math problemTorsten Slok, chief economist at Apollo Global Management, warns that the math in the AI space doesn’t add up yet. The massive profits at the top of the AI food chain are driven by investors, not customers, he said.“Capital can bridge the gap for a while, but not indefinitely,” Slok wrote in a report last week. “And therein lies the risk: Will the ROI show up for AI’s end customers fast enough to sustain the spending that is generating those upstream margins?”After analyzing profit margins in the S&P 500, Slok found no evidence that AI is boosting the profitability of healthcare, consumer staples, energy or real estate.However, Jeetu Patel, president and chief product officer at Cisco, said it’s hard to equate the AI revolution with the dotcom bubble because supply was built ahead of demand in the late 1990s.Patel, whose company is building infrastructure supporting the AI boom, expects AI agents to go mainstream, allowing demand to continue its explosive growth.“The demand is there (today), and supply is massively short on power, data center capacity, compute, memory and network,” Patel told CNN at Ai4, an industry conference in Las Vegas, this month. “We’re in the very, very early infancy.”AI is the dominant force in the economyThe stakes are massive – and not just for little-known hedge funds gambling on AI stocks. The US economy has become more reliant on AI and its enormous spending.The eye-popping gains of Nvidia, Micron, Alphabet and other AI stocks have helped inflate the net worths of millions of Americans by lifting the value of their retirement nest eggs, college savings plans and brokerage accounts.“Without the AI boom, we probably would be in a recession,” economist David Rosenberg said during a recent episode of the “Excess Returns” podcast.And timing is everything in booms and busts. As economist John Maynard Keynes famously said: “The market can remain irrational longer than you can remain insolvent.”Consider what happened to Julian Robertson, the legendary hedge fund investor.In the late ‘90s, Robertson correctly identified the massive dotcom bubble. But his bets against overvalued tech stocks blew up because the Nasdaq kept going higher and higher.Robertson eventually shut down his Tiger Management hedge fund in March 2000 – just as the Nasdaq was beginning what would become a historic crash.“Just because you think things are frothy doesn’t mean it’s time to get out,” Franklin Templeton’s Gokhman said. “The best returns occur when the party is about to end.”The-CNN-Wire™ & © 2026 Cable News Network, Inc., a Warner Bros. Discovery Company. All rights reserved.
Read Next Story