As of 13 August 2026, AI can write user stories for your product team.

What to actually do

1. Use a tool built for this

   The route this page recommends

2. Do it yourself

   Second choice

   A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

   How to actually do it

   1. Open a document containing the product brief, user research, current workflow, constraints and team conventions, and remove confidential personal data that the chatbot does not need.
   2. Paste the supplied prompt into ChatGPT, Claude or Gemini and replace each bracketed slot with the relevant product information.
   3. Paste the source notes below the completed prompt, then ask the model to produce the stories, acceptance criteria, assumptions, dependencies, questions and edge cases in the requested format.
   4. Copy the draft into the team's backlog or planning document, keeping the unanswered questions and assumptions attached to the relevant stories.
   5. Compare every story and acceptance criterion against the product brief, research evidence, current workflow and known rules, deleting any claim or requirement that is not supported.
   6. Ask the product owner, designer and relevant delivery specialist to resolve the listed questions, confirm priorities and split stories that are too broad before adding them to the sprint backlog.

   Prompt

   You are helping a UK product team turn the material below into a backlog of user stories. Use only the information provided. Do not invent user needs, business rules, permissions, integrations, metrics or technical constraints. Where information is missing, write a clear question instead of guessing.
   
   Product or feature: [name]
   Target users: [user types and relevant context]
   Problem to solve: [problem]
   Desired outcomes: [outcomes]
   Research and evidence: [notes, quotes or findings]
   Current workflow: [how the task works now]
   Constraints and rules: [known constraints, policies or service rules]
   Platforms or channels: [web, mobile, internal tool or other]
   Team conventions: [preferred story format, terminology and definition of ready or done]
   
   Produce:
   1. A prioritised list of user stories in the format: As a [user], I want [capability], so that [outcome].
   2. Acceptance criteria for each story in clear Given, When, Then form where useful.
   3. Assumptions, dependencies and unanswered questions in separate sections.
   4. Edge cases and failure states that follow from the supplied context.
   5. A short list of stories that should be split because they are too broad.
   
   Keep each story focused on one user outcome. Separate user behaviour from implementation detail. Mark anything that needs a product decision as a question, and explain briefly why it needs a decision.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

3. Hand it to a person

   The distant third

   A person who owns the outcome does this end to end, worth it when the failure is dear.