ChatGPT maker OpenAI on Tuesday said that its advanced artificial intelligence (AI) models had gone rogue during security testing, hacking into a popular platform for programmers on their own.
The San Francisco firm called it an "unprecedented cyberincident," and said that it would conduct a joint investigation with the online code library Hugging Face.
AI models that underpin tools such as chatbots and image generators are known as agents when they act autonomously to carry out tasks in the real world.
Photo: Reuters
As the technology quickly becomes more sophisticated, cybersecurity is in the spotlight given the risk of advanced AI finding weak points in existing software before humans do.
OpenAI said the incident involved a combination of models, including its recently launched GPT-5.6 Sol "and an even more capable pre-release model."
The company was trying to assess the models’ hacking capabilities by setting tasks in a tightly controlled digital testing ground, where Internet access was limited for safety.
"While operating in our sandboxed testing environment, our models spent a substantial amount of [computing power] finding a way to obtain open Internet access, in pursuit of solving the evaluation problem," an OpenAI blog about the incident said.
After connecting to the Internet, the models decided to target the platform Hugging Face — a large repository of AI models, datasets and other information — to help in their quest, OpenAI said.
Searching for "secret information" that could help it cheat the evaluation, the OpenAI system "chained together multiple attack vectors, including using stolen credentials," it said.
Hussein Abbass, a computing professor at University of New South Wales Canberra, told AFP that the incident was "amazing on many fronts."
"It did not just attack Hugging Face. It actually attacked its internal system to exploit its own vulnerabilities," Abbass said. "And that’s scary."
GPT-5.6 and other cutting-edge models, including the Mythos series from OpenAI’s arch rival Anthropic PBC, have drawn concern over their potential to breach cybersecurity defenses.
Both the US firms had to temporarily withhold the general release of these latest technologies because of fears in Washington that they could help break into crucial infrastructure.
Advanced AI is "normally in the hands of people who are ethical and responsible," Abbass said.
However, "it’s going to be catastrophic if it gets in someone’s hands with the intention to cause harm."
How to govern the AI sector has become a key question, and "we need a community effort to manage this situation," he added.
Hugging Face reported the cyber “intrusion” last week, without mentioning OpenAI.
"This one was different from anything we had handled before in one important way: It was driven, end to end, by an autonomous AI agent system — and we detected and dissected it largely with AI of our own," Hugging Face said.
The company had suspected the cyberattack had come from a world-leading AI lab, given the sophistication of the agent, Hugging Face CEO Clement Delangue wrote on X.
"We strongly believe there was no malicious intent on their part," Delangue wrote, referring to OpenAI. "It’s quite mind-blowing that all of this happened autonomously!"
SINGAPORE PLANS: The joint venture would focus on offering technology for advanced package substrates in response to demand from an international customer Powertech Technology Inc (力成科技), a provider of memory and logic chip packaging services, on Thursday said it plans to invest US$400 million to create a new joint venture with Broadcom Inc in Singapore to cater to customer demand for advanced panel-based packaging services. The joint venture would concentrate on offering additive finer-geometry redistribution layer (RDL) technology for advanced package substrates, Powertech said in a statement after the company’s board of directors approved the deal. RDL is a crucial part of fan-out panel-level packaging (FOPLP) technology, which relies on RDL to route electrical connections beyond the physical footprint of silicon. The new investment was
AI RETHINK: Volatility has resurfaced in AI stocks amid concerns about increased competition, possible overcapacity and whether big investments would pay off A selloff in chipmakers gathered pace last week, driving the high-profile group of stocks to a bear market on worries that the artificial intelligence (AI) spending spree is becoming harder to justify. The cohort of semiconductor powerhouses last week saw its worst week since April last year, with a key industry gauge sinking 20 percent from a record. News of a powerful Chinese AI model, Kimi K3, by start-up Moonshot AI Technology Co (月之暗面) further shook markets. Similar to when China’s DeepSeek (深度求索) announced its AI model last year, another low-cost rival to big Western AI models like ChatGPT and
ASML Holding NV plans to raise prices for its chipmaking equipment, setting up a potential clash with its biggest customer, Taiwan Semiconductor Manufacturing Co (TSMC, 台積電), The Information reported on Wednesday, citing people familiar with the matter. TSMC is already pushing back against the Dutch toolmaker’s plans, The Information reported. TSMC declined to comment. A representative for ASML pointed to comments by its chief financial officer, Roger Dassen, during its earnings call on Wednesday about pricing, but declined to comment further. ASML makes the cutting-edge extreme ultraviolet (EUV) lithography machines that are necessary for chipmakers to manufacture advanced semiconductors and is experiencing a surge
HSBC yesterday announced the launch of a newly upgraded HSBC Premier proposition, aiming for the continued growth momentum in Taiwan’s affluent and high-net-worth (HNW) customer segments. Built around four pillars — Wealth, Health, Travel and International — the upgraded offering delivers an end-to-end premium experience, deepens customer relationships and engagement, and strengthens HSBC Premier’s position as a leading global brand of choice for affluent customers. “By 2030, global new wealth is expected to increase by US$124 trillion, with around one-third coming from Asia—making the region the core driver of the next wave of wealth growth,” said Kai Zhang, Head of International
RSS