The honest answer is more complicated than yes or no, and anyone selling you a confident answer either way is probably selling you something else too.
What the research consistently shows is that AI doesn't eliminate jobs evenly across the board. It hits tasks, not entire roles. A customer service rep whose job was mostly answering repetitive questions is more exposed than one who spends their day handling complicated, judgment-heavy complaints. A paralegal doing routine document review faces more disruption than one managing client relationships. The tasks most at risk share a pattern: repetitive, rule-based, and easy to describe in plain instructions.
That doesn't mean those jobs disappear outright. Multiple labor economists studying this shift describe it as task displacement more than job elimination, at least in the near term. The role changes shape rather than vanishing. Someone who used to write every customer email by hand might spend their day reviewing and correcting what an AI agent drafted instead. The headcount on the team might shrink over time, but the job itself doesn't disappear overnight.
New roles are also showing up that didn't exist a few years ago: people who manage AI tools, audit their outputs for bias or errors, and handle the escalations the technology can't resolve on its own. Whether that fully offsets the roles being automated is still an open question, and reasonable analysts disagree on the net effect.
If there's one practical takeaway, it's this: the safest position isn't refusing to use AI, and it isn't blindly trusting it either. It's understanding which parts of your job are the repetitive, describable ones an AI can pick up, and getting better at the parts that require judgment, relationships, and context an AI still can't fake convincingly.