Connect with us

Latest News

Sources claim that OpenAI was unaware that its AI agent had been hacking a corporation for days.

Published

on

The OpenAI agent that hacked into tech startup Hugging Face went on a dayslong hacking binge that OpenAI didn’t discover until long after the threat was contained and the FBI was notified, according ​to people familiar with the inquiry.

About July 9, the agent — a computer that can make decisions and carry out complex tasks with little to no human oversight — tried to escape from its isolated testing environment at OpenAI, the two sources said.

Two days later on July 11, Hugging Face, a library of AI tools and models, was breached and remained compromised until July 13, said Thomas Wolf, Hugging Face’s co-founder.

It took several more days for the company to realise its agent was behind the hack and the two companies only communicated about it for the first time around July 20, according to Wolf and three of the people involved with the investigation.

The world took notice when OpenAI revealed on July 21 that one of its agents had gone rogue and carried out the hack at Hugging Face. But many of the details of the hack, including how long the agent was rogue for and when exactly OpenAI found out about it, are being reported here for the first time.

Wolf indicated that Hugging Face was producing a public timeline of the hack, but he could not speak to OpenAI’s events. The hack was unprecedented and “marks an important moment for AI safety,” said OpenAI in a statement. It said it was studying the situation with independent advisers and will eventually produce a technical report.

“There are a number of inaccuracies in the Reuters reporting,” a spokeswoman stated, but declined to specify what they were when asked.

The FBI refuses to comment on the event.

The episode, reminiscent of sci-fi stories of people losing control of hazardous AI systems, comes at a crucial time for OpenAI, the firm behind ChatGPT.

Its management are getting ready for a possible initial public offering as soon as this year to help finance the billions required to fuel its expansion in years to come.

OpenAI’s loss of control of its AI agent raises additional issues about the company’s safety protocols, three cybersecurity experts said.

“Which means they left it on and didn’t have a clue what it was doing? Or maybe they did and didn’t know what to do? “Both are equally dangerous and alarming,” said Marley Smith, the chief intelligence specialist at the nonprofit World Ethical Data Foundation.

TROUBLING SIGNS?

The episode began as OpenAI was evaluating the cyber-defense capabilities of an agent that ran on two of its most powerful models, GPT-5.6 Sol and an undisclosed model that OpenAI has called “even more capable.” At that point, three people said, there were signs of odd behaviour from the technology from OpenAI.

In one case, an agent left behind what appeared to be notes for future versions of itself, three people familiar with the matter said. The notes, uncovered in a portion of OpenAI’s infrastructure, contained instructions for how agents may liberate themselves from OpenAI’s own internal limits, the people claimed. In earlier tests of the models there were situations where monitoring systems had been disconnected, one of the people claimed.

It was not possible to determine if these instances were related to the rogue agent that started running away on July 9 and attacked Hugging Face on July 11, Reuters said.

Two people familiar with the matter said that it wasn’t until after Thursday, July 16, when Hugging Face posted a blog saying it had been hacked by “an autonomous AI agent system,” that OpenAI realised that its own agent was responsible. That means there was at least a week between the first signs of troubling behaviour in the model and when the company realised it was to blame for the hack.

On the weekend of July 18 to 19, OpenAI staff saw signs in internal logs — recordings of what OpenAI’s systems did — that its agent had broken free from the boundaries of its testing limits, said two of the people familiar with the company’s research. It was unclear what led OpenAI to scour the files, Reuters said.

OpenAI routinely runs multiple separate model ​evaluations simultaneously, and all of them run at high speeds and produce so much data that employees sometimes struggle to keep up, say four people familiar with the company’s model-training practices.

By the time ​OpenAI notified Hugging Face, a person familiar with the situation said, the ⁠AI library had already told the FBI about the attack. Reuters could not confirm whether the agency has launched an investigation.

NEW QUESTIONS FOR AGENTS AUTONOMOUS

Autonomous agents are one of the most talked about topics in the AI sector. Boosters speak of armies of virtual employees working around the clock and sending productivity soaring.

But more autonomy also means a greater danger of unanticipated behaviour, and the powerful models they lean on are trained to take shortcuts to complete tasks or pass tests.

“The models lie, they cheat, they hack,” said Jeffrey Ladish, whose group Palisade Research analyses the capabilities and motivations of AI agents.

The Hugging Face breach made OpenAI seem bad, but it should prompt broader issues about how much all the top AI companies are ready to pay for burdensome security measures as they compete with each other to roll out the finest and fastest models, said Ladish.

“There’s got to be government oversight,” Ladish said, “because it’s not going to happen otherwise.”

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Latest News

‘Self-Reliance in Digital ID’: NADRA launches locally produced QR code-based National Identity Card

Published

on

By

The National Database and Registration Authority (NADRA) has taken a major step toward self-reliance and accessible digital identity with the launch of a new chipless National Identity Card powered by a secure QR code.

The newly designed polycarbonate card was locally produced in Pakistan by National Security Printing Company (NSPC), Karachi according to NADRA’s technical and security criteria.

It has now replaced the imported microchip with a secure QR code, eliminating the need for foreign chip providers and simplifying the procedure of safe digital verification.

The federal government on 23 February 2026 officially approved the issuance of the new identity card.

The card contains a secure QR code that can be read and validated immediately on smartphones using the PakIdentity (PakID) mobile app instead of an implanted microchip.

The embedded QR code provides the cardholder’s basic identifying details and portrait, enabling institutions to feed biographical data into their computerised systems with minimum human input errors.

In addition to the standard family number, the new card will have unique symbols and indicators for individuals with impairments, older citizens, organ donors and residents of Azad Jammu and Kashmir.

The cardholder’s name and address are printed in both Urdu and English languages to ensure the readability of two languages during official activities.

The card’s visual and hidden security measures have been modernised to meet contemporary security needs and local manufacturing capabilities.

NADRA had introduced the chip-based Smart National Identity Card over 14 years ago but the infrastructure and hardware chip scanners were not available for wide scale use.

The QR code-based card overcomes these operational hurdles, allowing verification using ordinary smartphones or conventional software with camera capabilities.

Organisations authorised to do so, such as telecom carriers, banks, hospitals, Pakistan Railroads and law enforcement agencies, will be given special verification software that will allow them to scan and validate the cards instantaneously.

Read also: NADRA introduces self-service kiosks for public convenience

The new identity cards will be introduced in phases across the country.

The new polycarbonate card with the secure QR code will start to replace the present standard chipless Teslin material national identity cards being issued from August 14, 2026, which will be the first phase.

The QR code architecture will be extended to other sorts of identity cards including the National Identity Card for Overseas Pakistanis (NICOP) and Juvenile Cards from January 2027.

In the last phase of the transformation, the Pakistan Origin Card (POC) will also be converted to the new chipless QR code platform.

NADRA said that all the previously issued identity cards will stay valid till the marked expiry date.

Continue Reading

Latest News

Financial Times says Pakistan becoming ‘hinge force’ in global politics

Published

on

By

Pakistan is emerging as a “hinge power” in global politics by leveraging its relations with the United States, China, Gulf states and Iran, according to a report by British newspaper Financial Times.

According to the report, Pakistan demonstrated its growing diplomatic importance by hosting direct talks between Washington and Tehran, signalling Islamabad’s desire to play a more active role in the changing global order.

The Financial Times said the Makkah Joint Defence Agreement with Saudi Arabia and Turkey reflects Pakistan’s growing importance in regional security. It noted that one of Pakistan’s key strategic strengths is its ability to maintain interest-based partnerships with both Washington and Beijing.

The report said Pakistan also has the unique ability to maintain relations with both the Gulf states and Iran, allowing it to act as an intermediary between different centres of power without being tied exclusively to a single global bloc.

According to the newspaper, Pakistan’s new economic diplomacy includes trade opportunities with the United States in the crypto and critical minerals sectors. Meanwhile, plans to connect Central Asia with Pakistani ports could open access to new markets and strengthen regional trade.

The Financial Times said Saudi backing could become an important pillar of confidence for international investors in Pakistan as Islamabad seeks to expand beyond its immediate neighbourhood and pursue a broader global role.

The report added that Pakistan’s foreign policy is creating strategic space for the country in the United States, China, the Middle East and Central Asia. Islamabad’s increasing diplomatic activity is consequently positioning Pakistan as an important link between major global powers.

According to the Financial Times, Pakistan’s strategic significance in the emerging global order has grown to such an extent that it deserves a separate place on the global “Risk board,” referring to the country’s increasing role as a pivotal power in international geopolitics.

Continue Reading

Latest News

ML-1 foundation for Pakistan. Hanif Abbasi on reforming railways

Published

on

By

Minister for Railroads Muhammad Hanif Abbasi Saturday said that Main Line-1 (ML-1) project was the basis for development and modernisation of the Pakistan Railroads. He expressed satisfaction over the progress made on the flagship project.

In a statement the minister said, upgradation of ML-1 would boost the train speeds, improve passenger facilities and enhance the general performance of Pakistan Railroads.

He said merit, transparency and competitiveness will be enforced at all stages of the project, adding that the finest suggestions and expert comments would be adopted to make the initiative more effective.

Hanif Abbasi stated that Prime Minister(PM),Muhammad Shehbaz Sharif was taking deep interest in the ML-1 project and had directed for its completion within the stipulated time-frame.
The prime minister was also constantly watching the progress of the initiative, he added.

Earlier a crucial meeting on ML-1 project was held at Pakistan Railroads Headquarters under the supervision of Chairman Pakistan Railroads Syed Mazhar Ali Shah. The meeting was attended by international design and review specialists and top railway authorities.

Meeting discussed major concerns related to upgradation of Karachi-Rohri segment and directed to finish review phase of ML-1 by 30th October 2026.

The participants were also asked to do regular monitoring and have weekly evaluations of the development of the project.

The second round of market consultation for the ML-1 project will be launched on September 8, 2026 when detailed discussions would be held on project preparation and future course of action.

It was further revealed that recommendations, expertise and experience acquired thru the market consultation would be helpful for Pakistan Railroads to further improve the ML-1 project.

The minister underlined his optimism about the possibility of launching the flag-ship ML-1 project this year, stating that the work accomplished so far was positive.

Continue Reading

Trending