Artificial Intelligence | AI regulation under the direction of corporations

Anthropic boss Dario Amodei is a welcome guest at, among other things, the World Economic Forum in Davos. Photo: dpa/AP/Markus Schreiber Are it ethical scruples that are driving Anthropic boss Dario Amodei's call for slowing down and better control of the development of AI systems, or just his company's profit interests? In fact, the two can hardly be separated from each other. On the one hand, certain risks are worrying, but on the other hand, they are also likely to be bad for the image and could be expensive. And so Anthropic reacts as companies always do when there are obvious grievances: by proposing voluntary self-restrictions. To ensure that the competition in China does not rush away, export controls on AI chips should help. No wonder that our colleagues from Open-AI & Co. agree. Big Tech in the US could live better with such a slowdown than with the current situation. Especially since the proposed independent inspectors act anything but independently of the clients. Transparent AI research in public spaces would be an alternative to short-term profit-oriented developments.

It is probably not a coincidence that Amodei fails to mention that there have long been various agreements on artificial intelligence at the UN level. These have so far been lacking in commitment and the lack of sanctions for violations. The fact that AI agents are threatening humanity's existence in the near future is of course greatly exaggerated, as is the warning of a complete loss of control. But they can cause significant financial damage. And it would be necessary to hold the companies that developed these models liable, which the industry of course wants to avoid at all costs. One idea is for AI companies to pay into a fund that is intended to insure against damage. This could also be used to finance the capacity building called for by the UN Secretary General, including control systems, in the global south, which has so far been almost completely cut off from AI development.

It is clear that international regulation needs to evolve, with and not against China. The ultra-right government in the most important player, the USA, can't do anything with something like that. Even the weak suggestions from the industry go too far for her. As before, the free market should direct AI development so that domestic companies continue to expand their lead and China loses out. Whatever the cost.

Even the Big Tech bosses are much more progressive. But what is needed is truly independent, democratically legitimized regulation of the AI ​​industry. This will not succeed if it takes place under the direction of corporations.

The “nd.Genossenschaft” belongs to those who read and write it. With your contribution, you ensure that our journalism remains accessible to everyone - without any media company, billionaire or paywall.

→ report independently and critically → focus on overlooked topics → give marginalized voices a platform → counter misinformation → initiate and further develop left-wing debates

You can make the difference with “Voluntary Pay” or a cooperative share. You help keep this newspaper alive. So that nd remains.