Tags

, , , , , , , , , , , , , , , , , , ,

Recent weeks have seen an impressive burst of new information about how U.S.-owned businesses are fueling the technological and military strength of China, a country whose armed forces American soldiers, sailors, and airmen and women could be fighting on the battlefield before too long.

The first source of this information comes from Georgetown University’s Center for Security and Emerging Technology (CSET) in an October report called Harnessed Lightning: How the Chinese Military is Adopting Artificial Intelligence.

The study focuses on China’s own efforts to develop artificial intelligence (AI) capabilities and incorporate them into its military operations and systems, and goes into fascinating detail about how much money is spent on these efforts, and how many Chinese entities of all kinds are involved in the campaign. The authors also make clear – just in case it wasn’t screamingly obvious already – how widespread these applications can be, and their incredible potential to revolutionize warfare and hand victory to the power possessing the best knowhow.

But as one of the team explained in a summary magazine article two days ago:

Our research also highlights that U.S. companies are inadvertently powering Chinese military advances in AI. The overwhelming majority of advanced computer chips at the heart of China’s military AI systems are designed by U.S. firms like Intel, NVIDIA and Xilinx, and manufactured in Taiwan. We found that suppliers actually depicted NVIDIA-branded processors in photos of their products, providing clear evidence of the role U.S. technology plays in powering China’s advances. One company, which won a contract to supply chips for the PLA Strategic Support Force, even bought the domain ‘nvidiagpu.com.'”

Moreover, much more than simply semiconductors are involved. So is machine-learning and intelligent text-processing software, along with systems for “real-time monitoring” of “millions of global shipping and related users” with the help of 108 satellites from the American company Orbcomm.

My only objection: It’s inconceivable that these U.S. firms don’t fully understand the national security implications of their activities. The report itself notes that

Because most institutions that supply AI-related equipment are new and not subject to end-use controls, the Chinese military is frequently able to access or acquire technology from abroad, including from the United States. Some Chinese suppliers make a business out of sourcing foreign data or components and reselling them to sanctioned Chinese defense companies or PLA [People’s Liberation Army] units.”

But the U.S. businesses must be aware that any of their products sold to any Chinese entity are going to be made available to the Chinese military simply because that’s the way China has operated since the Communists have been running the place. So this rationale can be easily laughed off.

The same cynical reaction is justified for claims that U.S.-owned firms don’t know that the capital they’re steering into the Chinese tech sector will also benefit the Chinese military. And these capital flows are both impressive and coming both from finance companies and from the huge semiconductor manufacturer Intel – which is hoping to receive billions in U.S. government subsidies and tax breaks to help restore its competitiveness in microchip production largely (of course) to bolster national security.

As reported by The Wall Street Journal this morning, Intel is “is among the active investors, backing a Chinese company now called Primarius Technologies Co., which specializes in chip-design tools that U.S. companies currently lead in making.” RealityChek regulars, moreover, know that Intel has been investing in other defense-related Chinese entities for years.

Not that American investment firms aren’t also doing their part to strengthen China’s tech prowess and therefore military capability and potential. Including the Intel deal, the Journal found, American companies “participated in 58 investment deals in China’s semiconductor industry from 2017 through 2020, more than double the number from the prior four years….”

And on top of these transactions, according to the Journal, “the China-based affiliates of Silicon Valley venture firms Sequoia Capital, Lightspeed Venture Partners, Matrix Partners and Redpoint Ventures have made at least 67 investments in Chinese chip-sector companies since the start of 2020….” In all, the sums involve run into the billions.

And in case you still doubt that these U.S. firms fully understand how valuable their investments are to a country that’s increasingly hostile to America, the Journal article quotes the head of one of these Chinese recipients as saying that his operation is working with the Chinese regime and other partners “to help our country get rid of its dependence on foreign high-performance chips.” Since the United States is still ahead in this sector, a China that no longer relies on American high tech products is going to be a China that’s caught up – and possibly grabbed the lead.

What’s the U.S. government doing about this dangerously unacceptable situation? It’s true that Washington has long maintained a system of export controls aimed at preventing China and other worrisome countries access to critical, militarily relevant goods and knowhow. But as the CSET study documents, this system is being completely overwhelmed – in part because of sorely inadequate funding and staffing, and in part because it’s never switched from a case-by-case approach to the kind of much broader denial strategy that’s clearly needed for a systemic threat like that posed by China.

There’s legislation in the works to plug some of the holes, and according to the Journal, the Biden administration seems supportive. Let’s just hope that the government gets its act together sometime before weapons powered by American technology and funded by American investors start killing American servicemen and women somewhere in East Asia.

BTW, thanks to friend Bill Holstein for calling my attention to these two items.