Welcome everyone to another edition of my China Tech Law Newsletter. Today, I’d like to recommend readers check out a great new paper by Matt Sheehan for the Carnegie Endowment for International Peace called China’s AI Regulations and How They Get Made. Matt is also the author of an amazing book called The Transpacific Experiment: How China and California Collaborate and Compete for our Future, and has recently turned his attention to China’s emerging AI governance approach.
That approach is seen by some as a chance for China to try to stay ahead of the curve on the sophistication and changing nature of AI applications. Matt’s comment in the paper here seemed spot on - that developing a reputation for being a global leader on AI governance was simply a “nice to have.” More so than a genuine motivating factor for the urgency to regulate an emerging technology sooner rather than later.
When we talk about China, especially these days, we often have to acknowledge the different political systems that China’s regulations are a product of compared to the Western world. Even if there are major differences in motivations, there are still some ven diagram-like overlapping motivations for developing AI governance shared by China and the West. Such as wanting to minimize the “myriad social, ethical, and economic impacts AI is having on people in China”. The tension between the seemingly endless business applications and the corollary negative societal implications.
While the primary motivation, as Matt rightly points out, for China is control of published content, that doesn’t preclude an honest interest by regulators (who are real people btw) from looking at how AI governance can actually protect individuals from harmful influences.
We’ve seen the same in other contexts. For example, on data regulations. China developed the Cybersecurity Law, the Data Security Law, and the Personal Information Protection Law after GDPR came online in Europe but before many other countries have developed their own frameworks. The motivation was both control over data, which had national security and macro-economic importance, but at the same time genuinely trying to push back against commercial exploitation and lack of security for individuals' data. Another ven diagram.
Similarly, the recent anti-monopoly (or anti-tech monopoly) crackdown, even if exempting important actors like state-owned companies, still was attempting to break the stranglehold of a small duopoly of internet giants, Alibaba and Tencent, that did have too much market dominance for consumers’ own good. Again, there was a dual motivation – regaining government control over increasingly powerful private companies, but also a sincere desire to stop companies from limiting consumer choice. Just as Matt points out, an important feature of these AI regulations is to limit how algorithms can create anti-competitive business practices or excessive price discrimination, as well as protecting workers.
I further agree with Matt’s take that even if we know the primary motivation in these rules is often control, we can’t dismiss the value they have in other areas and a governance framework worth potentially borrowing from (even if piecemeal). There are more problems in common to address than there are differences.
As Matt says, people say regulations in China don’t matter because they can be arbitrarily ignored. But they do matter. Just ask any company familiar with the Chinese Labor Contract Law, now 15 years old and how it has radically altered workers' rights (often giving employees too much power). As I can attest to in dealing with many difficult employment situations over the years, all employees know the law inside and out and its words most definitely matter.
Things have come a long way as far as the sophistication and foresight of Chinese regulators on technology especially. As I wrote about for my experience 10+ years ago with deciphering encryption regulations, it always helps to have experience reading tea leaves for clients in the past to anticipate how a law will evolve next.
Concluding, I like how Matt points out that there is a tremendous amount of jockeying among government agencies in this new AI regulatory framework. Perhaps even much more so than in the US, there are major turf battles for control over policies at the operational level.
The Cybersecurity Administration of China (CAC) seems to have the upper hand in so much as Matt says it has been an inclusive facilitator getting buy in, but watch out for the Ministry of Industry and Information Technology (MIIT) as AI applications in industry become more and more prevalent. We got into that point and others in a recent interview I did with Paul S. Triolo as well.
Check out the interview with Paul, and most definitely check out the paper by Matt. Between the two, some great minds are helping us understand AI policy development in China. I really encourage all readers of this China Tech Law Newsletter to take a look.
And btw, I do appreciate the followers of this blog. We’re pushing close to 3,000 subscribers now, if you haven’t subscribed yet, well what are you waiting for?
And on that note, see you back here in a couple weeks!
*This blog may be considered attorney advertising. It is for informational purposes only and does not constitute legal advice.