As product managers, we already wear many hats - shaping strategy, delighting customers, aligning stakeholders, and delivering outcomes. Now, AI is adding a new dimension to the role.
At Atlassian, we see this shift firsthand. AI isn’t just a buzzword - it’s reshaping how PMs work. The truth is simple: AI won’t replace product managers, but PMs who use AI will replace those who don’t.
Adoption is already happening. HubSpot’s State of AI Report shows that 28% of professionals use AI to collect and analyze customer feedback - a core PM responsibility. AI aids writing and coordination. Teams use LLMs like ChatGPT to draft specifications, release notes, and marketing copy (e.g. idea brainstorming and outline planning). Specialized tools summarize communications: multiple tools ingest meeting transcripts and Slack chats to generate status reports, action items, and notes. And there are so many more use cases.
Whether you’re building AI features or not, you can’t afford to ignore this. The good news? It’s not too late, and it’s not difficult to use. In this write-up, I’ll share how we use AI at Atlassian - and how you can bring it into your own day-to-day PM or other craft work.
Why Building with AI Matters More Than Ever
We’re entering a new era where AI isn’t just a co-pilot - an active participant in how work gets done. As tools become more powerful and more alike, a lot of the work PMs do will get commoditized. The real differentiator won’t be which tool you use, but how you think and what you bring to the table that AI can’t.
AI already handles much of the heavy lifting: drafting, summarizing, translating, synthesizing. It can automate meeting notes, turn hours of user interviews into insights, or tidy up Slack updates - giving you back time for the work that actually moves the needle.
But here’s the key: AI can’t (well at least right now) do strategy. It can’t feel empathy. It can’t make product judgment calls. It can’t replace the sense-making and intuition that great PMs bring.
The PMs who thrive will be the ones who use AI not to replace their craft, but to amplify their uniquely human strengths.
Think of it as an investment in time. Every task you automate gives you space for:
- Sharper strategy
- Deeper customer conversations
- Better navigation of ambiguity
- Higher-quality decisions
Three years from now, two PMs may both have access to the same AI. What will set them apart is the quality of their thinking.
The AI Learning and Application Flywheel for PMs
Adopting AI as a product manager isn’t about hacks - it’s about building a muscle. A simple flywheel keeps you moving forward:
- Learn through small wins
- Apply by scaling what works
- Share to sharpen your own skills and help others
This cycle compounds over time, turning you into an AI-empowered PM.
Step 1: Learn with Small Wins
“Start with simple use cases like summaries and writing to build confidence. Just use the LLMs wherever possible.”
You can deliberately block your calendar for 30-60 minutes weekly to learn or plan to do at least one task every day using AI.
Examples:
- Meeting Summariser: Take your meeting scripts and summarize them, identify the actions and key inputs.
- User Interview Synthesis: Turn raw notes into themes and insights.
- Slack Update Rewriter: Convert messy updates into clear comms.
These small wins require no expertise - just curiosity.
Pro tip: PMs have limited time. Focus on learning LLM applications rather than technical details, especially when starting.
Step 2: Apply through Prompt Libraries, Agents & Prototyping
Prompt Libraries
A great prompt has structure: role, task, format, constraints, context (cannot stress how important context is).
At Atlassian, PMs contribute to a shared AI PM Playbook, crowdsourced with prompts for strategy, roadmapping, analysis, stakeholder management, prototyping, and more. Below is a quick glimpse into some things you as a PM can do with the help of LLMs and AI (many of these feature in Atlassian’s PM Playbook and have well defined prompts/Rovo agents that can be readily used internally).
Reusable Agents
When a prompt proves useful, codify it into an agent or a ChatGPT project. At Atlassian, we build Rovo Agents in order to reuse the prompt and get our day-to-day work supercharged. For example, I use PRD creation, PRD critique, and Slack message creation agents quite a lot. I have also tried experimenting with using agents to block meetings with cross time zones and multiple people and find the best slot - my agent needs more work but is getting there.
Prototyping with AI
Beyond writing and synthesis, AI is becoming a powerful prototyping partner. You don’t need to wait for a designer to get started:
- Generate wireframes from a simple product description
- Create user flows to validate journeys
- Draft mock content to test messaging early
- Simulate user scenarios to explore edge cases
At Atlassian, PMs often use (and are encouraged and now have access to approved 3P tools) AI to spin up rough prototypes in hours, which accelerates feedback loops and frees designers and engineers to focus on refinement instead of starting from scratch.
Step 3: Share to Accelerate Learning
The final step is to share what you learn. Every time a PM contributes a prompt, demo, or agent, the flywheel spins faster for everyone.
Ways to share:
- Help create a prompt library for your organisation or even a small team. Add to your team’s prompt library.
- Showcase your work or prompts in demo sessions.
- Document and share learning on LinkedIn and other common interest groups.
Sharing isn’t just generous - it’s how you get better too.
Why This Matters
Over time, every PM will have access to the same AI tools. What will separate good from great is how effectively they learn, apply, and share. The faster your flywheel spins, the more value you - and your team - create.
Final Thoughts: Seize the AI Advantage
The future of product management is undeniably AI-powered—and the leaders who thrive tomorrow will be the ones who start experimenting today.
Begin with the basics. The power of AI compounds over time, and the earlier you build the muscle, the stronger you’ll be. We’re already seeing companies build incredible businesses by leaning into AI. Take Gumloop for example: with only two full-time employees, this U.S.-based workflow automation startup raised a $17M Series A. Their ambition? To scale to a $1B valuation with just a 10-person team. That’s the kind of leverage AI makes possible—small teams doing work that once required hundreds.
Of course, stories like this are outliers—but they’re also signals of where the world is heading. The race is real. The exciting part? It’s fun. Using AI every day isn’t just about efficiency; it’s about discovery, creativity, and seeing ideas come to life faster than ever.
So don’t wait. Start learning, keep experimenting, and share your best PM tips and tricks for using AI in your work. The more we exchange what’s working, the stronger we all get.
Co-authored with Premanku Chakraborty. Originally published on LinkedIn, Oct 23, 2025.