← Latest papers
💻 computer science

A systematic review of generative AI usage for IT project management

Using the PRISMA methodology, this systematic review explores the current state of generative AI in IT project management, finding that research is currently in an exploratory stage dominated by GPT-based prompt engineering and suggesting future directions such as specialized AI agents and human-AI collaborative networks.

Original authors: Ionut Anghel, Tudor Cioara

Published 2026-04-27
📖 4 min read☕ Coffee break read

Original authors: Ionut Anghel, Tudor Cioara

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The "Digital Co-Pilot" Revolution: Making IT Projects Less Chaotic

Imagine you are trying to build a massive, complex LEGO castle with a team of 50 people. Everyone is talking at once, some people are missing pieces, the instructions are 500 pages long, and halfway through, someone decides the castle should actually be a spaceship. It’s a nightmare of communication, lost parts, and constant changes.

This is what IT Project Management often feels like. Software projects are huge, fast-moving, and incredibly complex. Traditionally, humans have had to manually track every single "brick" (task), every "instruction" (requirement), and every "builder" (developer).

This research paper explores how Generative AI (the brain behind tools like ChatGPT) is stepping in to act as a high-tech "Digital Co-Pilot" to help manage this chaos.


1. The Current State: The "Smart Assistant" Phase

Right now, we are in the early stages. Think of GenAI as a very bright, very fast intern.

If you ask this intern to read a 100-page manual and summarize the most important parts, they can do it in seconds. In IT projects, this "intern" is currently being used to:

  • Write the "Manuals": Automatically drafting project plans and documentation.
  • Organize the "To-Do List": Turning messy ideas into clear, organized tasks (like turning a brainstorm into a structured shopping list).
  • Check the Work: Looking at code to find bugs or spotting when a project is starting to drift off course.

However, the paper notes that we are mostly just "talking" to the AI (using something called prompt engineering). We haven't yet fully taught the AI how to "think" like a professional project manager.


2. The Future Vision: From "Intern" to "Expert Dream Team"

The researchers propose that we are moving away from just having one chatbot and moving toward a "Digital Dream Team" of AI Agents.

Instead of one general assistant, imagine having a specialized crew of digital experts working alongside your human team:

  • The Specialist Agents (The "Department Heads"): Imagine having a digital expert for every stage of the project. One agent is a master at Planning (the Architect), one is a master at Execution (the Foreman), and one is a master at Closing (the Auditor). They don't just chat; they actually manage specific parts of the workflow.
  • The Role-Based Agents (The "Virtual Teammates"): Imagine if, during a meeting, you had a "Virtual Scrum Master" or a "Virtual Risk Manager" sitting at the table. They wouldn't just take notes; they would speak up and say, "Hey, if we take this path, we might run out of budget in three weeks!"
  • The Hybrid Network (The "Orchestra"): This is the ultimate goal. It’s a symphony where all these specialized AI agents work together in a network. They talk to each other, share data, and prepare everything perfectly, while the Human Project Manager acts as the Conductor. The conductor doesn't play every instrument, but they make the final decisions, handle the "emotional" side of the team, and ensure the music sounds right.

3. The "Speed Bumps" on the Road

It sounds like magic, but there are three big reasons why we aren't there yet:

  1. The "Hallucination" Problem: Sometimes, an AI can be so confident that it tells you a lie (like saying a LEGO piece exists when it doesn't). In a multi-million dollar project, a "confident lie" can be a disaster.
  2. The "Privacy" Problem: If you tell an AI your company's secret plans, where does that information go? Keeping data safe is like making sure your digital assistant doesn't accidentally shout your bank password in a crowded room.
  3. The "Trust" Problem: Humans need to understand why an AI is making a suggestion. If an AI says, "Stop the project immediately," a manager won't listen unless the AI can explain its reasoning clearly.

The Bottom Line

The paper concludes that AI isn't coming to replace the Project Manager; it’s coming to upgrade them. We are moving from a world where managers spend all their time doing "paperwork" to a world where they spend their time doing "orchestration"—using a powerful digital crew to turn chaos into organized, successful creation.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →