Just like the tittle said, we should create laws that help prevent over usage of AI, making sure that risky AI development like for a driverless vehicle is properly enforced and if it is too risky then it is prohibited, and promote the development of none risky AI. An example of this is the EU Artificial Intelligence Act that rate AI into 4 risk level: unacceptable, high, limited, and minimal or no risk. Depending a level, the usage or development of a particular AI may be prohibited, limited or freely implemented.

Another method is to educate everyone about not just the advantages of AI about the advantages of AI but also the harm it could cause, and ways to prevent it. IF everyone is well educated about the harm and usefulness of AI especially those that are not tech savvy, then they can make a better choice of not developing risky AI system; prevent the idea of replacing every " human things" with AI; or not supporting a particular AI usage as even though it is useful, it causes too much harm; etc... An example of this is what this website is doing right now, educating and helping the reader understand the usefulness, harm, and and way to prevent harm from AI. 

Another method is to speak out against companies or government that are over relying on replacing human worker with AI. For the past years, many companies in the movie industry, art industry, game industry has decided to try to replace their actors, voice actors, artists, etc... with GenAI as a way for them to increase speed and efficiency, in their word. And so this lead to the SAG-AFTRA strike 2 years ago against the movie industry where actors, voice actors, artists, etc...  fight to prevent AI taking over their jobs. And in the movie industry, they have won but their fight is still going on with the gaming industry.