Simple: ban authors and co-authors for x years when caught.

Also don’t publish a paper with even one hallucinated citation.

Soon I think people will learn to work around it backwards, retrospectively writing the non-existent papers. This way you get a brand new paper in your CV with a citation count already above zero.

Why the ** don't institutions announce zero-tolerance towards fraud, and hire lawyers to proactively sue and claw back earnings from fraudulent employees?

Why don't communities literally go out and do citizen's arrests on academics who commit fraud? Get rid of them professionally and replace them with more honest, better motivated people