AI companies like Anthropic seems to be driving the narrative that AI is starting to get too powerful and they need governments to step in and create regulations before it’s too late.
The skeptics on the other side claims the AI companies are pushing for these regulations because they don’t want to be held accountable for the actions of their own models.
The skeptics think the existing laws that are in place already incentivises the AI companies to focus on safety and engineering. Because if the AI does start to cause damage the companies that developed the AI would be held liable for damages.
A good example for this is Tesla. Tesla is incentivised to make their self driving as safe as possible because if the car get into a crash because of the models not making the right decisions they would be held liable
Similarly if your local claude code in the computer starts cyber attack then Anthrophic would be held liable. There is no need for any additional regulations.
A badly written regulation could take away liability from the AI labs.
Plus the regulation could even make it harder for open source models to be developed and used by individuals. Given the open source models are starting to get better and cheaper, it’s also in the best interests of these huge ai companies to make it harder to build and release models that you can run locally.