Newsletters

Both the American and Chinese governments have signalled they plan to rein in open-weight AI models. 

The Trump administration seems certain to add some Chinese technology companies to the Entity List to protect investment in American frontier AI models. Meanwhile, Beijing is apparently weighing applying export restrictions on Chinese AI tech, including open-weight models.

Overnight, different US government officials issued a strong, coordinated signal that they plan to take action against Chinese AI companies. Michael Kratsios, the director of the White House's Office of Science and Technology Policy, wrote on X that Chinese company Moonshot AI had "developed a sophisticated internal platform to conduct large scale distillation against US models" and that "large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable."

The Linux kernel project has disclosed 442 vulnerabilities over the past three days, in a massive dump of CVEs on its security mailing list.

Although not confirmed, the bugs were likely discovered using AI tools. Over the past months, projects like Anthropic's Glasswing and OpenAI's Daybreak have been granting access to advanced frontier cybersecurity models to top-tier security firms and researchers to find bugs with AI in major open-source projects.

Most of the bugs are low-severity issues, so nothing world-ending for the internet today.

A hacker has breached Romania's cadastre agency and wiped the country's entire land registry database following a failed extortion attempt.

The hack has brought Romania's entire real-estate market to a standstill as official apps and websites have been offline for a week. Notaries can't record new transactions while citizens can't obtain proof of ownership or detailed land records.

Email servers at the National Agency for Cadastre and Real Estate Advertising (Agenția Națională de Cadastru și Publicitate Imobiliară, or ANCPI) were also down as part of the incident.

It's commonly thought that ransomware operators are simply using AI to hack more companies. However, some operators are employing AI to generate leverage so they can extract more money in negotiations with victims. 

A leading example is FulcrumSec, a data extortion group that started operating around September 2025 and uses simple techniques to breach organisations. It typically gains access by taking advantage of hardcoded or exposed credentials, unpatched applications or misconfigured storage. 

The group claims to have breached 25 organisations and stolen several terabytes of data using these techniques. 

The Indian government has ordered Apple and Google to remove a battery management app from their stores that had been used over the past weeks to hack e-rickshaws across the country as part of a toxic viral trend that mocked drivers and risked blocking traffic.

Videos of the hacked have been spreading on social media for two-three weeks in a trend named the Tirri Challenge.

The videos showed distressed drivers stuck in traffic pushing their vehicles to the side or crying over lost revenue on the phone with family and mechanics.

A new US Supreme Court decision could undermine the flow of Section 702 intelligence collection from Europe.

Section 702 allows the US government to collect intelligence on people outside the US with compelled help from communication service providers and is regarded as the crown jewel of American foreign intelligence collection. 

Since 2000, successive data sharing agreements have facilitated transatlantic commerce by allowing the legal transfer of personal data from the EU to the US. Section 702 collection from Europe relies on these data flows and the intelligence program has been at the heart of legal challenges to the data sharing agreements. 

All new cars manufactured and sold in the EU and in the US will have to include a mandatory infrared camera aimed at the driver's face at all times, which is alarming some privacy groups.

The infrared camera will track the driver's head position and eye movements and then alert distracted drivers if their eyes go off the road.

The new regulation has entered into effect in the EU on Monday and will enter into effect next year in the US. In the EU, the new camera requirement is part of the block's second General Safety Regulation (GSR2), a broader swath of new safety rules introduced for the auto industry and designed to improve road safety.

Android 17, released last month, has shipped with stricter protections against lockscreen PIN and password guessing attacks.

Google has reduced the maximum number of failed attempts from 1,800 to just 20, and the timeouts between failed attempts are now way more aggressive.

The previous Android 16 allowed ten wrong guesses in the first minute, which increased up to 1,800 across five years.

The developers of a lot of industrial gear and smart devices will have their work cut out for them over the coming months and years to deploy protections against a set of newly discovered and unpatched bugs in the FatFs filesystem driver.

The seven bugs, discovered by security firm runZero, can allow an attacker to use a crafted filesystem image to cause a memory corruption that runs malicious code to jailbreak a targeted device.

Devices that use FatFs for their filesystem are all impacted.

Last week Anthropic accused Chinese company Alibaba of conducting what it described as the "largest known distillation attack" against the company's AI models. 

Distillation attacks upskill less capable models by training them on the outputs of more advanced ones. Back in February Google, OpenAI and Anthropic all said that Chinese companies were harvesting their proprietary intellectual property in coordinated campaigns.

Alibaba's latest campaign, Anthropic says, occurred from April 22 to June 5 and used more than 25,000 fraudulent accounts to generate 28.8 million exchanges. Anthropic says it was carried out by operators "affiliated with Alibaba and Alibaba Qwen, Alibaba's AI lab".