

No disrespect to you, but it is clear you don’t understand the technology well.
Algorithms are, per definition, routine computations to perform a task. Even clicking on a link and then the browser requesting and displaying the page associated with it is an algorithm!
And about (generative) AI: -They cannot be made energy efficient because generation (giving you a text or image) is incredibly energy consuming -You cannot build an AI without bias or discrimination, because all your training material has to be unbiased and without discrimination - and social progress means that we will continue to recognize previously unknown harms in society. AI can only be as unbiased as its most biased training material. And the amount of data you need for training is too high to be efficient to gather before training. -AI cannot be made to be accurate. It is definitionally impossible, because it is basically an extension of auto-correct - it can only predict the next likely word from a huge collection of data. Accuracy is never a part of the algorithm of AI. -In terms of public service, it would be better to point out above chatbots that they can only generate legibile text, all other aspects are secondary.
So, in summary, what you suggest is technologically impossible, or at least misguided.
Shooting yourself in the foot so others can’t shoot you in the foot? No.
How about outlawing and banning the shooting of feet instead? That might work better.