Basically if you go back a few years ago "AI" was basically marketing speak only. Engineers and researchers said "ML" almost exclusively. So it was a good litmus test to identify people who actually knew what they were talking about vs hype people

Then AI entered the public lexicon with GPT becoming popular. So people shifted to saying AI because everyone was confused when saying ML. Still "ML" is used commonly in researcher to researcher conversations (depending on the niche, but I think the flag is still a good indicator here)