Peer reviewers are fellow academics who are not paid. Peer review does not add any cost to publishing. The cost all goes to for-profit companies making record profits.
Peer review absolutely add costs. Finding a reviewer who's qualified, free and willing takes time and effort, and that's increasing by the day. Frankly, I think peer review is doomed. There's now a flood of AI-generated papers that journals need to weed out, and even 99% success is not good enough for peer-reviewers' workload and continued respect. On top of that, natural demographics, the cuts to university funding in the West and the improvement of education in Africa, Asia and South America mean there is a greater and greater proportion of young researchers to experienced, overburdened researchers, and that's just too much load to bear.
As soon as you pay reviewers, you now have to deal with people pretending to be academics and pretending to review a paper just to earn money, which adds additional cost. Not to mention that money can introduce biases even if the reviewer is genuine.
Depending on the discipline, there are plenty of people with PhDs who are adjuncting for low wages. And for other disciplines, plenty of people with PhDs who work in industry. And plenty of people without PhDs who know enough to find errors in papers (I've found conclusion breaking errors in two so far).
