First, doctors get busy in hospitals, and in offices, and they can and do miss things. This is like another pair of eyes looking at data. The AI has no ability to decide what is wrong, order anything, or prescribe treatment. It's a flag. Not sure how you find that problematic. This has been started to be used with radiologists as well. Not to take work away, but have someone look behind them, or maybe point out scans that might deserve more time.
Now, will this flag be aimed to increase profits? Might be. Certainly now, plenty of doctors do things to either increase their revenue or decrease costs. They avoid some tests, or they might do something like add procedures. Lots of evidence that doctors and hospitals make decisions for profit. Lots more cesarean procedures, lots of other little profit generating procedures that take little time and make $$$. Ai is no different. Humans will change things to benefit themselves, which is why we ought to have more transparency on how humans and AIs make decisions.