I'm a doctor and I've been following this saga for a while. What you wrote does not match my experience. Firstly, you overestimate our political reach by a large margin. If you can have acceptable service for massively cheaper, it will happen regardless of any lobbying. What _is_ true, is that AI systems outperform docs for very select, often trivial situations (e.g. routine chest X-ray). I don't believe this would shrink the market for radiologists significantly. Secondly, the non-trivial work is exceedingly difficult to automate, because those cases currently have a prevalence to complexity ratio that make them impossible to train for. The US are making great strides towards that kind of thing, though. But not available yet.