Are you ready for co-creation? And, why this is the most important skill shift for many on project teams.
Most conversations about AI are built around the same question: how can AI help do work faster? It is a reasonable question. It is also the wrong starting point.
Speed is a byproduct of working well with AI. It is not the purpose. And when speed becomes the frame, we optimize for the wrong things. We use AI to produce outputs more quickly while doing the same kind of work in the same kind of way. We feel more productive. We may actually be falling behind. And, leaders are starting to see that actual gains are difficult to find.
For most of the history of software development, teams and each role has had reasonably clear edges. The roles have been designed to make coding efficient and less expensive. The work of each role had a shape and a container. Agile ways of working have tried to dissolve these containers and focus on the work and skills over titles on the team, yet few organizations have been successful with the organizational design needed to truly achieve agility.
AI is being heavily used in software development as the initial agentic systems organizations are using. The agility, process, roles, and work design is changing quickly and the role containers are dissolving. AI fluency, agentic ai work design, and decision making are the new bottlenecks.
The work teams are being asked to do now does not stay inside traditional boundaries. It merges into territory that belongs to many roles. It forces us to collaborate and co-create in ways that most have never experienced. Needs analysis, process design, systems design, user experience design, data flow analysis, and solution architecture is work that is simultaneously happening as teams build, increment, and test faster than ever. Yet, with incremental and iterative precision, this co-creation still needs skills, structured knowledge, and discipline to get a sustainable pace and quality results. For example: A business analyst or product owner facilitating a rapid prototyping session with business leaders and the development team is doing work that sits at the intersection of discovery, design, and validation all at once.
AI is part of what makes this blurring possible. It extends what one person, or one team, can think through and produce. But it only creates that expanded capability when we use it for expanded work, not just to compress the time it takes to do the old work.
The task has moved. The way we use AI needs to move with it.
Using AI With Others, Not Just On Your Own
The most underexplored use of AI in business analysis is collaborative AI. Using it not as a solo productivity tool but as a shared thinking resource in the room with others.
Think about what that looks like in practice.
A BA sitting with a business stakeholder, product owner and a developer, using AI in real time to stress-test assumptions about a business process with a real working, clickable prototype. Someone asks a question the team has been dancing around. The BA prompts the AI to model what would happen if a particular assumption were wrong. The group sees it, reacts, and the conversation shifts. That is not the BA using AI to produce a document faster. That is the BA using AI to accelerate shared understanding in a way that would have taken days through the traditional back-and-forth of text and documents.
Or a BA facilitating a requirements session with a stakeholder, using AI to generate a rapid draft of a process flow mid-conversation. The stakeholder can see something and react to it. “That’s not quite right, the exception happens here, not there.” The BA adjusts. The AI adjusts. The stakeholder is no longer being asked to react to a blank page or a document written in absentia. They are in the work with you.
Or a BA working alongside a developer to analyze an existing process for AI integration. Both people are contributing. AI is synthesizing, questioning, filling in gaps, and generating options for the team to evaluate. The BA is not handing off a requirements document. The BA and the developer are co-creating the design together, with AI as the third participant extending both of their capabilities.
This is collaboration at a new level. And business analysts, whose core skill has always been bringing people into shared understanding, are extraordinarily well positioned to lead it. But it requires us to think differently about what our role in these sessions is. We are not just the person who writes things down afterward. We are the person who knows how to use AI to accelerate what the group is thinking through together.
AI Should Make Us Smarter, Not Just Faster
There is a version of AI-assisted BA work that feels like progress and mostly is not. It is the version where AI writes your requirements, summarizes your meeting notes, and formats your process flows. Outputs appear faster. The BA reviews them, adjusts them, and ships them.
That is fine. It is not the transformation.
The version that actually changes what BAs can do is the one where AI makes the thinking better. Where it surfaces connections the BA would have missed. Where it asks a question that reveals an assumption nobody had examined. Where it holds complexity that would have exceeded working memory and allows the team to see the whole picture at once.
AI is extraordinarily good at pattern recognition across large amounts of information. It is good at generating options quickly. It is good at stress-testing logic. It is good at working across domains simultaneously in a way that human cognition struggles to sustain. These are not speed capabilities. They are thinking capabilities. And when a BA uses them well, the work gets better, not just faster.
The shift is from using AI to produce to using AI to think. That sounds simple. In practice, it requires a different relationship with the tool and a different quality of engagement.
The Cognitive Switch Worth Paying Attention To
This is something that does not get talked about enough: working with AI the way described above requires a gear shift that most of us have to make deliberately.
The default mode is production. Prompt, receive, review, adjust, ship. It is efficient. It feels like using AI well. And it is a mode that most of us slide into naturally because it resembles how we have always worked, just faster.
The thinking mode is different. In thinking mode, you are not using AI to generate something you already know you need. You are using AI to explore something you do not fully understand yet. You are asking it to challenge you, to extend you, to surface what you are not seeing. You are engaging critically with what it produces, pushing back, redirecting, and building something through the exchange that neither you nor the AI would have produced alone.
That mode takes more from you. It requires presence, skepticism, and active judgment throughout. It is cognitively demanding in a way that production mode is not. Many BAs, when they try to work this way, find that they need breaks more often than they expected. The quality of focus required is different.
Developing awareness of which mode you are in is a real and valuable skill. Not because production mode is wrong, but because if you are always in production mode, you are leaving the most important capability on the table. Pay attention to the difference between how it feels to use AI for output and how it feels to use AI for thinking. They feel different. Learn to recognize both.
What Working Differently Actually Looks Like
If this is the direction, what does it mean concretely?
It means bringing AI into collaborative sessions, not just solo work. It means being willing to use AI in front of stakeholders and teammates, not just behind the scenes. It means developing the facilitation skill of knowing when to invoke AI in a group session and how to use the output to move the conversation forward rather than derail it.
It means using AI to generate questions as often as you use it to generate answers. When you are stuck on a complex problem, asking AI “what questions should I be asking here that I am not asking?” is often more valuable than asking it for solutions.
It means treating AI output as a first draft of thinking, not a finished product. Always. The moment you begin to accept what AI produces without genuine critical engagement, you have stopped using AI to think and started using it to avoid thinking.
And it means building reflection into how you work. After a session where you used AI collaboratively, what landed? What did AI surface that the group would not have gotten to on its own? Where did it miss the mark and why? This closes the loop and makes you better at this kind of work over time.
The Work We Are Being Asked to Do
The business analyst role is expanding. The task has moved beyond its traditional edges, the work is increasingly collaborative and hybrid, and AI is the tool that makes the new scope of work possible.
But AI only delivers on that promise if we use it for what it is actually good at: making the thinking better, bringing teams into shared understanding faster, and expanding what we can collectively hold and work through.
The BAs who will lead in this moment are not the ones who use AI to move faster through the old work. They are the ones who use AI to do new work, together, that the profession has not been asked to do before.
That is the shift worth making.