07/01/2026
A pattern we keep running into: Jira admins and the Atlassian users with configuration access (project admins, folks spinning up Rovo agents, etc.) are leaning on AI to make design decisions, and it's quietly steering them toward setups they'll regret.
🤖 The trouble is rarely that AI gives a wrong answer. It gives a reasonable one. Ask it how to handle a specific request and it returns a clean, by-the-book solution that aligns with best practice in general. What it can't account for is the context that actually matters: how your teams really work, what's already built, and how that tidy answer holds up at scale. It solves the question as asked. What it misses is the need underneath, the thing an experienced person would have caught before the question was even finished.
🔧 We use Claude every day, but always with an experienced person guiding it, directing it, and correcting it when it drifts. That's the part that matters. AI is a genuinely useful tool in the hands of someone who knows the terrain and can see the second and third order effects. Handed the keys and trusted to be right, it becomes the source of the mess instead of the fix.
The companies that get the most out of AI aren't the ones using it the most. They're the ones who pair it with people who ask the right questions first, then know when to say "not like that."
