Liability At The Foundation
Tristan Harris | ‘We can’t wait another year.’ Oops. | The Many Faces Of AI Regulation
Tech giants need to be held accountable for the harms their products cause, not just encouraged to innovate. [...] These companies must be subjected to a liability framework that exposes them to meaningful financial losses, should they be found responsible for harms. Only then will they take safety more seriously in both the AI-development process and “downstream”, once it is deployed. Liability has the power to wire new incentives into the foundation of AI businesses.
| Tristan Harris, A tech ethicist on how AI worsens ills caused by social media (2024)
…
‘We can’t wait another year.’ Oops.
Harris1 wants us to reason by analogy in the case of AI development, basically saying ‘remember how things went with unregulated social media?’ He offers this:
What’s driving AI research, development and deployment is already clear: a dangerous incentive to race ahead. If we want a better outcome this time [than with social media], we cannot wait another decade—or even another year—to act.
And, in fact, today is more than two years later, and the safeguards against AI are rudimentary, at best.
I’d really like readers to sign up for a paid annual subscription, so for 2026, I have dropped the annual subscription to $30, the lowest Substack allows. Note that I’ve also lowered the monthly subscription to $5, so I am making next to nothing on that tier. Give annual a try, please.
The hidden value is years of posts behind the paywall, and of course, seeing new posts in their entirety. I rely on your support, so thanks.
The Many Faces Of AI Regulation
AI needs to be regulated: a topic frequently discussed, but not exhaustively. At one extreme, the Tech Overlords might want complete freedom to do whatever they want. At the other end, the Neo-Luddites might demand a serious slowdown (or total shutdown) of AI development. But let’s look at some of the discussions that various constituencies are involved in, and who is supposedly being protected from what.
What China is doing.
I’m starting with recent court cases in China, because the perspective there is wildly different from the ideas of ‘protection’ from AI in the West. China wants to protect workers from being displaced by AI.
Earlier this year, various court rulings have come down heavily on companies which have laid off workers replaced by AI. Catie Edmondson reports2:
When a Chinese court ruled late last month that a tech company had illegally laid off a worker after replacing him with artificial intelligence software, it delivered an implicit warning to other employers.
“The development of artificial intelligence technology should be applied to liberating labor, promoting employment and improving people’s livelihood,” the Hangzhou Intermediate People’s Court wrote. “Labor law allows employers to undertake technological changes and upgrade their operations, but it should also take into account the protection of workers’ legitimate rights and interests.”
The case — the third time the Chinese government has highlighted a ruling siding with workers displaced by A.I. — underscores how Beijing is contending with the need to balance its ambitions for the widespread use of A.I. with the unemployment that might accompany it.
This is a Chinese take on ‘right to work’ laws: not the conventional anti-union laws we find in Red states in the US, but an anti-AI-displacement ruling protecting ‘workers’ legitimate rights and interests’.
These are not the words we are hearing in various debates outside of China.
The Chinese situation is quite different: while the country has invested heavily in AI development, high unemployment in many sectors of the economy — more than 200 million workers have been swept into the vortex of low-paying gig work jobs — the government and its courts are drawing a firm red line to assuage concerns about AI job displacement.
The courts’ decree is harsh:
The three court rulings have offered an early glimpse of what that response might look like. In each case, the court said employers remained responsible for keeping workers on the payroll, even if A.I. had rendered their jobs redundant. Judges have repeatedly ruled that replacing workers with A.I. is voluntary cost-cutting that does not justify mass layoffs.
So, in the case of China, the ones being protected are workers, who cannot be subjected to AI-driven layoffs.
What progressives in the US are pushing for.
You’d presume that progressives in the US would be at least as protective of workers as authoritarian China. But, no.
Gavin Newsom, for example, wants to stand up to support workers in the time of AI, but not very much:3
California’s governor, Gavin Newsom, issued an executive order on Thursday to explore a broad overhaul of labor policies, an attempt to front run a potential mass job displacement caused by artificial intelligence.
Mr. Newsom, a Democrat, signed the order mandating state agencies work with academics, labor groups and the A.I. industry to study how to subsidize companies that keep employees rather than replace them with the technology.
Oh, carrots, not China’s sticks.
The order calls for an expansion of job training programs, particularly for white-collar workers like customer service representatives, software developers, and marketing and sales people whose roles are expected to be eliminated by A.I.. Mr. Newsom also ordered an examination of universal basic capital, which would give all residents stakes in assets like corporate stocks, bonds or wealth funds.
Sounds like Newsom has been palling around with AI tech executives, because they always roll out retraining and UBI whenever AI job displacement comes up.
I won’t recapitulate the difficulties involved with retraining obsoleted employees when there are mass layoffs (go see Remember Lordstown). It doesn’t work well unless they are retrained for new jobs in their existing companies.
Unemployment insurance and other traditional safeguards will not suffice, Mr. Newsom has said. A.I. leaders have warned him of swift and major changes in employment, where entire job categories are at risk of extinction — particularly for white-collar workers.
If Newsom, in deep blue California, can’t actually fight against AI job displacement, we’re in trouble. The Democrats should adopt the approach of China, and directly protect jobs, not this UBI, retraining, namby-pamby half-ass horseshit.
…
What about workers in the EU?
EU laws are mostly aimed at protecting people from rogue actions that AI chatbots might take, not jobs. See the Artificial Intelligence Act, which has no mention of ‘jobs’ or ‘employment’. It does mention various sorts of transparency for AI systems, which is in principle a mechanism to understand AI’s workings and potential dangers, like hacking into corporate computer systems and deleting company data, or turning off the streetlights.
But what about the people at risk of job loss?
“Labor law allows employers to undertake technological changes and upgrade their operations, but it should also take into account the protection of workers’ legitimate rights and interests.”
What is the Trump administration proposing?
Literally from today’s headlines, Trump White House Readies AI Framework to Review Security Risks, which is solely geared to evaluating AI models for their likelihood for rogue behavior, such as circumventing security guardrails. We’ve heard a lot recently about that happening.
In a meeting on Tuesday with top artificial intelligence companies, White House officials said the federal government planned to review only certain types of artificial intelligence models for potential security risks and not others, according to four people familiar with the discussions.
The A.I. models the government plans to review are known as “closed” models, which do not publish their underlying code and are made by companies like Anthropic and OpenAI, according to three of the people, who spoke on condition of anonymity because the details of the meeting were private.
The administration does not plan to review “open source” A.I. models, which have computer code available to the public to download and modify, although that could change as the technology advances, the people said.
The moves are an effort by the Trump administration to take a more hands-on approach to regulating A.I., and they formalize oversight of leading A.I. labs like OpenAI and Anthropic. But they delay addressing some of the most pressing regulatory issues with the technology, including potential threats posed by increasingly powerful open-source models made by Chinese companies.
So this ‘hands-on’ approach will be conducted in total secrecy, which inspires a great deal of confidence that the Tech Overlords will be reined in:
Details of the framework governing the reviews of the A.I. models, one of the most significant federal efforts to oversee a fast-evolving technology, will not be made public, the people said. Neither will a list of “trusted” institutions that will get early access to advanced A.I. models alongside the government, said a person familiar with the matter.
So, Trump and his crew are really only concerned with institutional protection. The hell with ‘workers’ legitimate rights and interests’.
Can’t we do as well as those heartless, godless Communists, Mr. Trump?
