Over the past two years, virtually every major project and task management platform has embedded artificial intelligence into its core offering. Asana launched its AI Studio and smart workflow capabilities. Atlassian rolled out Rovo, its AI-powered knowledge and coordination layer spanning Jira and Confluence. Monday.com positioned itself as an "AI work platform." Many more followed suit, each promising intelligent task assignment, automated status reporting, risk flagging, and natural language project planning.
Gartner analysts have projected that 80 per cent of today's project management tasks will be eliminated by AI by 2030 - a figure that has been cited in countless vendor pitch decks and flashy demos. The race to embed AI features has been less about demonstrated outcomes than about competitive positioning: not having an AI story is, increasingly, to look like you don't have a strategy, or you’re behind the times.
Do AI Project Management Tools Actually Boost Productivity?
For certain tasks and under certain conditions, yes. Whether AI truly ‘transforms’ how teams collaborate and deliver work is more complicated.
A 2025 systematic literature review published in MDPI found that AI delivers measurable gains in structured, data-heavy PM functions - schedule forecasting, resource allocation, risk identification, and earned value analysis. A peer-reviewed study in ScienceDirect mapped the effectiveness of AI tools across every project management knowledge area, finding that machine learning models outperform human estimators for timeline forecasting when high-quality historical data is available, and that generative AI meaningfully reduces administrative burden in communication and stakeholder management.
Generative AI has been demonstrated to provide modest productivity gains for individuals. The Federal Reserve Bank of St. Louis found that generative AI users saved approximately 5.4 per cent of their working hours - translating to a 1.1 per cent aggregate productivity increase.
The gap between individual gains and organizational results is where the story gets more complicated…
Asana's 2025 research found that 62% percent say AI outputs routinely fail to meet organizational standards, requiring additional review cycles. Fifty-five percent have had to completely redo work that AI generated. Furthermore, Atlassian's 2025 AI Collaboration Report found 96 per cent of companies have not seen dramatic transformational improvements from AI, despite workers reporting an average 33 per cent individual productivity uplift.
Organizations focused on AI-enabled coordination - rather than individual task speed - are nearly twice as likely to achieve organization-wide efficiency gains.
What Is Preventing AI From Delivering Results?
The barriers to effective AI adoption in project management are layered and compound. The ScienceDirect PM tools study found that 70 percent of practitioners lack sufficient understanding of which AI tools to deploy for which tasks, 62 percent cannot identify the most suitable application for their needs, and 58 percent cite inadequate technical infrastructure as a significant obstacle.
In other words, most organizations are deploying tools they don't fully understand, on systems that aren't ready to support them.
Even where understanding and infrastructure exist, a structural problem remains. Atlassian's data shows that 74 percent of workers feel held back because AI cannot access the right organizational data, and one in three knowledge workers admits to using unapproved tools to compensate, deepening the silos AI was supposed to eliminate.
Given the recent wave of mass layoffs attributed to AI, it is likely that employees are also somewhat hesitant to automate parts of their own workload and suspicious of agentic technologies.
After all, who wants to train a machine to do the job that pays their rent?
What Is AI-Washing and Why Should Buyers Be Cautious?
The collapse of London-based startup Builder.ai offers an illuminating case study. Founded in 2016, the London-based startup was one of tech’s most coveted unicorns. By May 2023, the company had reached a $1.5 billion valuation on the promise of an AI platform that could automate software development. In 2025, the Potemkin village collapsed when it was uncovered that their ‘neural network’ artificial intelligence platform wasn’t, in fact, artificial intelligence at all, but instead was predominantly composed of 700 Indian engineers. Builder.ai is not alone. The SEC charged Delphia and Global Predictions in 2024 for claiming AI-driven investment analysis that simply did not exist.
The recent Allbirds AI pivot, in which a failing shoe brand announced a strategic shift to AI compute in April 2026 and watched its stock price surge, captures the absurdity of the market’s logic.
With investors jumping at everything labelled AI, it is perhaps unsurprising that some of these claims are unfounded. As The Guardian recently reported, many companies are desperately positioning themselves as AI specialists to capitalize on the hype.
