How Program Managers Stay Valuable
ADVISORY ARTICLE
By Ankur Gupta
Senior Staff Technical Program Manager
LinkedIn
California, USA
Introduction: AI raises the bar for program management
For most of my career, program managers have been easy to recognize. We were the people carrying the plan, taking notes, chasing updates, maintaining the risk register, and turning a week of conversations into a Friday status report.
All of that work still matters. Programs still need accurate plans, clear records, timely updates, and disciplined follow-through. The people doing it were solving real coordination problems, often through substantial manual effort. What is changing is not the importance of the work, but how much time and effort it requires.
AI is making that kind of coordination dramatically cheaper. It can summarize a meeting, draft an update, compare plans, organize risks, and turn scattered information into a readable brief. Work that once occupied an afternoon can now take minutes.
This is usually framed as a threat to the program manager. I see it differently. AI is forcing us to separate the work that kept us busy from the work that made us valuable.
Once the administrative layer becomes easier, the question is not whether the program manager still has a job. The question is what the job was really for.
AI makes coordination easier, not optional
Nobody on my programs reports to me. I cannot offer a raise, grant a promotion, or rely on a reporting line to make ten teams move together. Yet I may still be responsible for getting a vice president to make a decision, carrying an architect’s direction into rooms they will never attend, and keeping several organizations pointed at one outcome.
That works because a program manager quietly holds a small part of many other people’s responsibilities.
The executive has the authority to align teams but not the time to sit in every working session. The architect owns the technical direction but cannot explain the reasoning in every adjacent meeting. Each engineering manager owns a team’s delivery but may not see the complete chain of dependencies around it.
The program manager connects those incomplete views. We carry context, detect when two teams believe different versions of the same decision, surface resistance that never appears in a status update, and create the conversation that allows work to continue.
This is different from moving information. Moving information is producing the minutes. Judgment is noticing that the polite agreement recorded in those minutes is not a real commitment.
AI is already good at the first part. It is not accountable for the second.
More…
To read entire article, click here
How to cite this paper: Gupta, A. (2026). When AI Handles the Work: How Program Managers Stay Valuable; PM World Journal, Vol. XV, Issue IX, September. Available online at https://pmworldjournal.com/wp-content/uploads/2026/09/pmwj168-Sep2026-Gupta-When-AI-Handles-the-Work.pdf
About the Author
Ankur Gupta
California, USA
Ankur Gupta is a Senior Staff Technical Program Manager working in AI infrastructure at LinkedIn. He has more than 15 years of experience leading complex, cross-functional technology programs across machine learning infrastructure, distributed systems, reliability engineering, and large-scale platform transformation. He is an IEEE Senior Member, published author, and speaker on AI infrastructure, engineering execution, and the evolving role of technical program management.
Ankur is based in San Jose, California, USA. He can be contacted via LinkedIn at linkedin.com/in/agupta03 and found online at ankurgupta.me.
