How AI Becomes a Political Crisis

AI is no longer distant tech news; it is becoming a political issue that politicians must address. The reason is not just jobless numbers, but that AI is hitting the career path young people value most: white-collar jobs.

1. 1% unemployment, political earthquake?

A 10% unemployment rate is not needed to shake politics. Just one industry or region suddenly hit by AI—say, mass accountant layoffs—makes front-page news. Everyone knows an accountant; such stories resonate with the public.

Worse, AI replaces entry-level white-collar workers. People can tolerate a 55-year-old steelworker taking early retirement, but if young people study hard yet cannot find work, that is a blow to every family's dream, with far worse political consequences than imagined.

2. Policy toolbox: what are the options?

Some seemingly clever policies may backfire. For example, taxing AI use would hit firms that adopt AI aggressively while protecting laggards, worsening market distortions.

In contrast, raising the corporate income tax is a more viable choice: it does not target specific companies, provides stable funding, and can be used to subsidize affected groups. But directly giving the government equity in AI companies, making it a shareholder, is a bad idea—it invites regulatory capture, biasing policy toward a few giants.

Interestingly, subsidizing entry-level jobs for 22-to-27-year-olds might be the most effective intervention. This prevents firms from collectively skipping recruitment, which would cost a generation their growth opportunities. Although it may be a "bridge to nowhere," it is better than doing nothing.

3. Lack of data, policy failure

A key problem: the government has no idea what AI is actually doing. Labs hold the best data but do not share it. Governments need to cooperate with labs to obtain API usage and task automation data to craft targeted policy.

Otherwise, policy is just groping in the dark.