> Why isn't it illegal to use traditional algorithms that have bias? Why aren't there regulations ensuring bias testing when humans themselves make decisions?
It is and there are, but rule-based systems and humans have traditionally been more limited.
You'd better bet that health insurers continually get audited in terms of claim pay rates and correctness.
The major difference is that AI systems can be pre-audited at minimal cost. In contrast, human auditing traditionally consumes billable processing time and so is done post-hoc on production results.
Honestly, I'm excited about these types of regulations. I think systems with more transparency, testability, and accountability are going to be better for society as a whole.
If it took "cost savings" + "scary AI" to get us there (digitization/automation and proactive outcome auditing respectively), that is what it is.
It is and there are, but rule-based systems and humans have traditionally been more limited.
You'd better bet that health insurers continually get audited in terms of claim pay rates and correctness.
The major difference is that AI systems can be pre-audited at minimal cost. In contrast, human auditing traditionally consumes billable processing time and so is done post-hoc on production results.
Honestly, I'm excited about these types of regulations. I think systems with more transparency, testability, and accountability are going to be better for society as a whole.
If it took "cost savings" + "scary AI" to get us there (digitization/automation and proactive outcome auditing respectively), that is what it is.