Ars Technica

Multiple Chinese APTs establish major beachheads inside sensitive infrastructure

View non-AMP version at arstechnica.com

A motherboard has been photoshopped to include a Chinese flag.

Hacking teams working for the Chinese government are intent on burrowing into the farthest reaches of sensitive infrastructure, much of it belonging to the US, and establishing permanent presences there if possible. In the past two years, they have scored some wins that could seriously threaten national security.

If that wasn’t clear before, three reports released in the past week make it abundantly so. In one published by security firm Kaspersky, researchers detailed a suite of advanced spying tools used over the past two years by one group to establish a “permanent channel for data exfiltration” inside industrial infrastructure. A second report published Sunday by The New York Times said that a different group working for the Chinese government had hidden malware that could cause disruptions deep inside the critical infrastructure used by US military bases around the world. Those reports came nine days after Microsoft revealed a breach of email accounts belonging to 25 of its cloud customers, including the Departments of State and Commerce.

The operations appear to be coming from separate departments inside the Chinese government and targeting different parts of US and European infrastructure. The first group, tracked under the name Zirconium, is out to steal data from the targets it infects. A different group, known as Volt Typhoon, according to the NYT, aims to gain the long-term ability to cause disruptions inside US bases, possibly for use in the event of an armed conflict. In both cases, the groups are endeavoring to create permanent beachheads where they can surreptitiously set up shop.

APT seeks long-term relationship with air-gapped device

A report published by Kaspersky two weeks ago (part 1) and Monday (part 2) detailed 15 implants that give Zirconium an entire gamut of advanced capabilities. The implants' capabilities range from stage one, persistent remote access to hacked machines, to a second stage that gathers data from those machines—and any air-gapped devices they connect to—to a third stage used to upload pilfered data to Zirconium-controlled command servers.

Zirconium is a hacking group that works for the People’s Republic of China. The unit has traditionally targeted a wide range of industrial and information entities, including those in government, financial, aerospace and defense organizations and businesses in the technology, construction, engineering, telecommunications, media, and insurance industries. Zirconium, which is also tracked under the names APt31 and Judgement Panda, is an example of an APT (advanced persistent threat), a unit that hacks for, on behalf of, or as part of a nation-state.

When I last covered Zirconium in 2021, the government of France had warned the group had compromised large numbers of home and office routers for use as anonymity-providing relay boxes for performing stealth reconnaissance and attacks. France’s National Agency for Information Systems Security (ANSSI) warned national businesses and organizations at the time that the “large intrusion campaign [was] impacting numerous French entities.”

The Kaspersky report shows that around the same time as the large-scale router attack, Zirconium was busy with yet another major undertaking—one that involved using the 15 implants to ferret sensitive information fortified deep inside targeted networks. The malware typically gets installed in what are known as DLL hijackings. These types of attacks find ways to inject malicious code into the DLL files that make various Windows processes work. The malware covered its tracks by using the RC4 algorithm to encrypt data until just prior to being injected.

A worm component of the malware, Kaspersky said, can infect removable drives that, when plugged into an air-gapped device, locate sensitive data stored there and copy it. When plugged back into an Internet-connected machine, the infected disk device writes it there.

“Throughout the investigation, Kaspersky's researchers observed the threat actors' deliberate efforts to evade detection and analysis,” Kaspersky wrote. “They achieved this by concealing the payload in encrypted form within separate binary data files and embedding malicious code in the memory of legitimate applications through DLL hijacking and a chain of memory injections.”

Disrupting US critical infrastructure

The series of attacks reported by the NYT came from Volt Typhoon, a group that hacks critical infrastructure and likely works on behalf of the People's Liberation Army. In May, Microsoft said it had “moderate confidence” that Volt Typhoon was “pursuing development of capabilities that could disrupt critical communications infrastructure between the United States and Asia region during future crises.” The software company said Volt Typhoon had specifically targeted critical infrastructure in Guam, where the US has military bases.

Citing American military, intelligence, and national security officials who mostly spoke on condition of anonymity, the NYT said that the Biden administration believes hackers related to the same group have planted malicious code deep inside networks controlling power grids, communications systems, and water facilities serving military bases in the US and elsewhere. The NYT continued:

The malware, one congressional official said, was essentially “a ticking time bomb” that could give China the power to interrupt or slow American military deployments or resupply operations by cutting off power, water and communications to US military bases. But its impact could be far broader, because that same infrastructure often supplies the houses and businesses of ordinary Americans, according to US officials.

The first public hints of the malware campaign began to emerge in late May, when Microsoft said it had detected mysterious computer code in telecommunications systems in Guam, the Pacific island with a vast American air base, and elsewhere in the United States. But that turned out to be only the narrow slice of the problem that Microsoft could see through its networks.

More than a dozen US officials and industry experts said in interviews over the past two months that the Chinese effort goes far beyond telecommunications systems and predated the May report by at least a year. They said the US government’s effort to hunt down the code, and eradicate it, has been underway for some time. Most spoke on the condition of anonymity to discuss confidential and in some cases classified assessments.

They say the investigations so far show the Chinese effort appears more widespread—in the United States and at American facilities abroad—than they had initially realized. But officials acknowledge that they do not know the full extent of the code’s presence in networks around the world, partly because it is so well hidden.

The discovery of the malware has touched off a series of Situation Room meetings in the White House in recent months, as senior officials from the National Security Council, the Pentagon, the Homeland Security Department and the nation’s spy agencies attempt to understand the scope of the problem and plot a response.

The recent Chinese penetrations have been enormously difficult to detect. The sophistication of the attacks limits how much the implanted software is communicating with Beijing, making it difficult to discover. Many hacks are discovered when experts track information being extracted out of a network, or unauthorized accesses are made. But this malware can lay dormant for long periods of time.

Mystery breach of 25 Microsoft cloud customers

The reports from Kaspersky and the NYT come on the heels of yet another big win by Chinese government hackers. In mid-July, Microsoft disclosed a mysterious breach involving its Azure and Exchange cloud services by yet another Chinese APT, this one tracked as Storm-0558. Through means Microsoft has yet to explain, Storm-0558 members acquired an inactive signing key used to grant access to Microsoft consumer cloud accounts.

After acquiring the highly sensitive key, the hackers somehow managed to use it to forge tokens for authenticating enterprise accounts on Azure AD. The supposedly fortified cloud service, in effect, stores the keys that thousands of organizations use to manage logins for accounts on both their internal networks and cloud-based ones. For roughly a month, the hack allowed Storm-0558 to track the email accounts of about 25 organizations, including the US Departments of State and Commerce and other sensitive organizations.

China has called the reports propaganda.

Post updated to correct the countries affected by Zirconium malware. The US wasn't included in the Kaspersky report.