You were hired based on merit. Regardless of the degree of AI contribution to your interview, you were judged and accepted based on your work, not your agents. You demonstrated your ability to your interviewers. You gained sufficient trust from your interviewers for them to recommend you as a potential coworker.
Agents did not gain that trust. They did not get hired. They did not get judged on their motivation or adaptability before being given the ability to work with you. Large language models work only within the context that they are given. Even with context they have occasional failures1. They think in a way unlike any human. They write in a distinctive, inhuman style. They cannot work fully independently2. These shortcomings can sometimes but not always be overcome3.
Your contribution is still required to achieve the best results. You cannot trust that your agent's work is fully accurate. Your agent will often lack the correct context to make the appropriate decision. Your coworkers want to know what you think, not just what your agent thinks. Your coworkers expect that you were a part of the work you produce. Representing your agent's work as your own is a betrayal of that trust.
Current tools often allow AI to act on our behalf. Be careful with this ability. It is easy to miscommunicate authorship. If you think the AI’s output is good enough to share without restating, mark it clearly as AI generated. Keep in mind that summarizing and restating the output of an agent will almost always result in clearer communication. Your coworkers are expecting you to use AI to enhance your work, not replace it. Your thoughts are more valuable than your agents. Share them.