AI’s speed paradox: why faster individuals don’t mean faster organizations

---

The learnings in this blog post are based on sessions from Atlassian’s Teamwork in an AI era event featuring Forrester Senior Analyst Will McKeon-White and Atlassian’s Teamwork Lab. You can check out these sessions and others on demand.

---

Everything’s moving faster. Decks are drafted in an afternoon. Customer emails write themselves. Reports that used to take a week now show up in your inbox overnight.

But the faster your teams move, the harder it gets to keep everyone on the same page.

At Atlassian, we recently surveyed 12,000 knowledge workers and over 170 Fortune 1000 executives for our forthcoming State of Teams report, building on findings from our AI Collaboration Index.

The good news? We found that 89% of executives say AI has made execution faster.

Unfortunately, half of those same leaders say cross-team coordination hasn’t improved at all.

So what’s happening?

AI can make it easier to get work done, but harder to work together

Leaders know the goal isn’t “make each person faster.” When we asked where they focus their AI strategy, only a small minority said at the individual level. Many more said they want to drive value for the organization as a whole.

But 58% of executives say they don’t know how to measure AI ROI at the company level, and nearly half say their high-level AI objectives are unclear. So when it comes time to show progress, they default to what’s easy to measure: individual productivity and efficiency.

The actual questions should be:

• Are we making better decisions as a company?
• Are teams moving in the same direction?
• Is it any easier to work across functions?

For many organizations, the honest answer to all three is still “not really.”

Why speed without coordination backfires

According to a Harvard Business Review study, about 80% of enterprise work is collaborative. By focusing almost exclusively on individual AI gains, companies may be leaving four-fifths of AI’s real potential on the table.

70% of knowledge workers say their processes and workflows are not optimized for AI, and 87% of teams say there’s no time or capacity for coordination because everyone is stuck in execution mode.

That means most of what matters in your business depends on people working together across roles, functions, and tools. You might write your part of the plan alone, but you only ship anything when marketing, product, legal, sales, finance, and operations all touch some piece of it.

Now, let’s add AI.

Suddenly, one person can crank out three versions of a proposal instead of one. A support agent can move through twice as many tickets. A marketer can draft ten campaign variants in a single morning.

Individually, these are wins. But collectively, they can become a problem.

Teams simply can’t absorb that much extra output if coordination does not improve. So what happens is what many of our respondents described:

• People moving faster in opposite directions
• Duplicative work
• “Final” decisions that don’t get broadly shared

And we are only at the beginning of this complexity curve. The rise of AI agents adds another layer of coordination tax. Work now has to be coordinated not just across people, but across people and agents.

Say a designer works with a design agent, a PM builds an agent to summarize customer feedback, and a sales team uses an agent to prep pitches. Each of those agents is helpful in isolation.

But if they’re not operating with the same context, they can easily recreate the uncoordinated chaos we often see in human systems, only faster.

We also found that executives predict structural shifts: 62% say AI is reducing silos, and 77% say teams will become more horizontal.

While reducing silos is good, it removes the artificial barriers that kept coordination costs low. In a silo, you only coordinate with your boss. In a blended, horizontal AI-enabled organization, you need to coordinate with everyone.

The missing ingredient: organizational knowledge

When we talk to teams about why AI hasn’t yet transformed their work, one theme keeps coming up: missing context.

Without business knowledge, AI just guesses.

Most organizations don’t have a single, shared understanding of basic things like what “good” looks like for a project, how a particular term is defined across teams, and why was a decision made three-quarters ago.

Instead, mission-critical context often lives in meetings that are never summarized, one-off conversations, people’s heads, and disconnected tools.

Lack of shared, easily accessible context has as high cost. Our research shows that knowledge workers spend hours each week in meetings where decisions are made, trade‑offs are discussed, and priorities are re‑set. When the meeting ends, that information rarely lands in a system of record. It just lingers in people’s memories and scattered docs.

Humans work around this by building informal networks. You know who to ping when the official documentation is wrong, you remember that “revenue” means one thing to finance and another to sales, and you have a mental model of which leader will quietly veto a decision if they are not consulted.

AI does not have that luxury.

If your organizational knowledge is fragmented, out of date, or locked away in places AI can’t access, then AI will behave like a very confident new hire: fluent, fast, and frequently off‑base.

The technology is not the problem. The knowledge layer is.

What the top teams do differently

The good news is that some teams are already breaking out of this pattern.

According to our research, teams that adopt a system for teamwork see a 68% reduction in their “fragmentation tax,” or how difficult it feels to work together. These teams were also nine times more likely to say AI helps them collaborate and three times more likely to fully trust AI’s outputs.

Top teams are not using a secret model or a magic prompt. Here are the three things they consistently do.

1. Start with purpose

Top teams are explicit about why they are adopting AI. They don’t just tell individuals to use AI, they identify their top challenges and then experiment to find ways for AI to help them solve those problems.

That clarity gives humans a reason to care that goes beyond “this is the new tool” and it gives AI something concrete to optimize for, instead of chasing generic productivity. AI can’t help you achieve your goals if it doesn’t know what they are.

2. Invest in documented context

Top teams treat organizational knowledge as infrastructure, not an afterthought. They put in the work to build a connected data layer. That means establishing a consistent place where decisions live, documenting clear owners for key concepts and metrics, and integrating tools so that AI can access a broad reach of documents, tickets, designs, and chats.

At Atlassian, our answer to this problem is the Atlassian Teamwork Graph: a connected data layer that ties together work, knowledge, and people across tools such as Confluence, Jira, Trello, Google Drive, SharePoint, Figma, and Slack.

The specific implementation matters less than the principle: AI can’t help if it can’t see.

When knowledge is shared and connected, AI can ground its answers in real context so teams can save time, make smarter decisions, and deliver better results.

3. Create a culture of learning

Finally, top teams treat AI as a long‑term practice, not a one‑time rollout. They rely on internal champions (the curious engineer, the operations lead who loves process, the support manager who has been quietly building macros for years) to experiment with early use cases, share what does and does not work, and then help colleagues translate abstract AI potential into concrete workflows

Leaders in these organizations invite the people closest to the work to shape how AI shows up. The sales rep who knows which tasks really eat up their day. The customer support agent who sees where tickets repeatedly get stuck. The designer who knows exactly which handoff meetings never need to happen again.

When you pair that bottom‑up insight with clear purpose and shared knowledge, you get a system where AI becomes embedded in how the team gets work done together.

What Forrester is seeing: high adoption, mixed impact

If this all sounds familiar, you’re not alone.

According to Forrester’s Digital Workplace and Employee Technology report, 90% reported using a generative AI platform by 2025, and 77% expected their organization to keep expanding that investment.

The impact, however, is mixed.

Will McKeon-White, Sr. Analyst at Forrester, noted that he sees roughly half of clients he talks to are seeing meaningful success. The other half are struggling – not only with adoption, but with the outcomes they had hoped for. Many have stories like these:

• An IT team that saves 18 hours a month for admins
• A media organization that cuts password reset time from five days to 15 minutes
• A sales team that improves conversion rates within a few months

These are real wins, but they’re not organization-wide transformations.

The shift from line items to transformation starts when leaders widen their lens. Instead of asking , “What can we automate?” they also ask, “What could we make space for?”

Their goal tends to be to give people back the time and attention they need to think, to experiment, to connect dots across teams.

From faster execution to better thinking

So, where does this leave leaders who are trying to make sense of it all?

McKeon-White recommends leaders ask themselves:

This thought experiment that forces you to look beyond time saved and instead consider:

• What would we do with more cognitive and emotional bandwidth?
• Which relationships would we deepen?
• Which bets would we finally have time to place?

Employees are most engaged and motivated when they can do meaningful work. At it’s best, AI creates that space. That is the real promise: not just more output, but more room for actual thinking.

Bringing it all together

If you feel like you are living the AI paradox – faster individuals, but a largely unchanged organization – start having your teams do three things differently:

• Start with purpose, not tools
• Invest in shared, connected knowledge
• Nurture a culture of learning, with internal champions and team‑led ideas

Then shift from asking “How much can we automate?” to “How much more meaningful, coordinated work can we make possible?”

The sooner organizations treat alignment, shared knowledge, and learning as core parts of their AI strategy, the sooner all that speed will start taking them somewhere worth going.

Learn how to adapt the way your teams use AI to create more meaningful, connected work. Check out sessions from Atlassian’s Teamwork Lab, Forrester’s Will McKeon-White, and other industry leaders in Teamwork in an AI era.