Zombie Projects at work are killing productivity. AI can put them to rest

We’ve all been there, returning from holiday to find work stalled, decisions deferred, or projects stuck somewhere between active and cancelled. At Atlassian, we call them Zombie Projects, and these skeletons in the work closet do more than linger in a Jira backlog or Confluence page — they drain time and attention.

They’re more common (and costly) than most realize. Our research shows that 44% of workers entered 2026 weighed down by Zombie Projects, and more than 90% say they’ve caused problems.

While leaders push for AI acceleration, with nearly 9 in 10 stating their organizations must move faster than ever to keep pace, 31% of teams say resources get wasted because of Zombie Projects. Halted work consumes the same capacity needed for meaningful experimentation and implementation.

Many teams talk about AI maturity while dragging unresolved work behind them. The impact is human just as much as operational, as 37% report feeling stressed, 32% report productivity loss, and 32% associate Zombie Projects with burnout risk.

In fast-moving organizations, unresolved work becomes a liability. So why don’t teams simply kill these projects?

Zombie Projects linger because of unclear ownership, with over 1/3 of workers fearing negative perception if they shut something down. Modern systems of work make it easier to leave tasks open rather than close them — avoidance that compounds into decision debt.

The result? Teams carry context they no longer use, burnout rises because inactive work still consumes cognitive load, and leaders operate with distorted views of capacity and priority.

The good news is that AI can help teams surface, evaluate, and resolve Zombie Projects before they accumulate into systemic drag.

AI as Zombie Project-fighting infrastructure

If Zombie Projects are a symptom of decision debt, AI can offer clarity. 60% believe AI could help guide revive-or-retire decisions, support they badly need. Our research found Fortune 500 companies waste 2.4 billion hours annually searching for information, much of it spent reconstructing context around work that may no longer matter.

AI doesn’t create time. But it can expose where time is quietly being wasted, especially within dormant work. Here’s how we use it in practice at Atlassian to quell the Zombie Project problem.

Start by defining communication standards

AI can only expose Zombie Projects if the underlying work signals are clear. One practical strategy we recommend is defining communication standards before projects kick off.

This looks like noting clear ownership on every initiative, defining status definitions (ie, what qualifies as in-progress), sharing regular updates tied to outcomes, and documenting decisions in shared systems.

Without this, stalled work can appear like active work. A project with vague updates and no milestones may look alive in a Confluence dashboard but be dormant in reality. By tightening communication norms, AI can more effectively detect inactivity, misalignment, and drift.

Squashing zombie projects up front with Project Posters

Another effective way to prevent Zombie Projects is to make intent explicit from the beginning. At Atlassian, we use Project Posters from our Team Playbook. These are living planning documents that clarify problems being solved, assumptions being made, outcomes intended, and measures of success.

Unlike static project charters, Project Posters evolve through conversations and force teams to articulate whether they’re solving the right problem rather than just shipping a solution. This matters for prevention because stalled projects often suffer from one of three issues -

• The original problem was misdiagnosed
• Assumptions were never validated
• Stakeholders were never fully aligned

By documenting intent, teams create a benchmark that AI can monitor project activity against and highlight where work is drifting. Instead of debating whether a project feels important, teams can assess it against its declared purpose, making structured closure more defensible.

Project Posters don’t eliminate Zombies on their own. But combined with AI, they make drift visible earlier, helping teams act before projects build up.

Continuous detection via agent partnership

Prevention is only half the equation because teams also need ongoing detection. At Atlassian, we use our AI teammate Rovo to help identify what is and isn’t top work.

Rather than manually auditing boards, agents can help summarize the most important initiatives, flag projects with declining activity, and compare active work against current priorities.

Additionally, instead of a leader scanning dozens of initiatives to determine which ones truly matter, agents can generate a ranked snapshot of work that’s driving measurable progress, work that’s consuming capacity without clear outcomes, and work that hasn’t materially advanced.

This shortlist becomes the revive-or-retire agenda, leading to greater efficiency and governance. When we surface misaligned and dormant work continuously, Zombie Projects have fewer places to hide.

Closing the Zombie Project gap

Zombie Projects result when teams tolerate ambiguity longer than closure. The good news is that pairing AI with decision clarity and discipline makes it possible to move faster and more confidently.

AI helps increase visibility, reduce ambiguity, and support clearer decision-making. It can continuously surface stalled work, flag misalignment, and rank tasks by impact so teams can distinguish progress from effort without results.

Project documentation clarifies purpose, assumptions, and goals, helping teams assess if work should continue or retire. Clear ownership, status definitions, and structured updates further keep AI insights reliable so teams can confidently revive or close projects instead of letting them linger.

When these practices are combined, teams can reclaim time, reduce cognitive load, and reallocate resources to say goodbye to Zombie Projects once and for all.