A smarter way to understand which Atlassian Cloud changes need your attention

[#1]

Hi Atlassian Community,

One of the most common things we hear from admins is that not every Atlassian Cloud change needs the same level of attention—but it can be difficult to know which ones do without reading them all.

Since launching the beta of Release Notes on Atlassian Community, we’ve heard from customers that having release notes in one place is a big step forward. We’ve also learned that teams need a clearer way to prioritize the changes most relevant to them.

Some updates are routine improvements. Others may affect permissions, APIs, user workflows, or how teams prepare for rollout. Today, it can still take significant time to read through each release note and determine what matters most.

So, based on early beta feedback, we’re working on a new AI-powered improvement to help make release notes easier to scan and prioritize.

Scan AI-generated impact ratings in the release notes feed, or filter by High, Medium, or Low impact to prioritize your review.

---

Introducing AI-generated impact ratings

We’re adding AI-generated impact ratings to help you quickly understand the likely significance of each release note.

Instead of treating every release note the same, impact ratings give admins and teams a faster way to identify which changes may need closer review.

Impact ratings can help you spot changes that may be:

• Low impact, such as routine fixes or small improvements

• Medium impact, such as updates that may require light awareness or validation

• High impact, such as changes that may affect workflows, user experience, permissions, APIs, or internal communications—for example, a change to how app permissions are granted

Each rating reflects the likely significance of a change based on the information available in the release note. It provides a starting point for prioritization, but it may not reflect the specific impact on your organization.

The goal is simple: help you spend less time reviewing every update with the same level of attention and more time focusing on the changes that matter most to your organization.

Impact ratings are meant to support prioritization, especially when you’re reviewing many release notes at once. You should still open the full release note to understand the details, rollout timing, and any actions your team may need to take.

---

A quick note on AI and beta

The new impact ratings are generated by AI. They’re intended to provide a helpful signal, not to replace your team’s own review or judgment.

Because this experience is still in beta, the ratings may not be perfect on day one. We’re starting with a lightweight version so customers can try it, tell us what feels useful, and help us improve the quality over time.

Your feedback will be especially important here. If an impact rating feels too high or too low, we want to hear about it so we can continue improving the experience.

---

Built to support change management

A simple place to start is by filtering for High and Medium impact changes. Review the full release note and rollout schedule, then decide whether your team needs to validate the change, communicate it internally, or take no further action.

Our goal is to make Release Notes on Atlassian Community a more practical starting point for understanding and managing Atlassian Cloud changes, whether you’re checking in weekly, planning for a release window, or deciding which updates need closer review.

---

Try it out

You can visit Release Notes on Atlassian Community here:

https://community.atlassian.com/release-notes

AI-generated impact ratings is rolling out as part of the beta. If you try them, let us know in the comments: Did the rating match your expectations? Was anything rated too high or too low? Your input will help us make the ratings more useful, accurate, and actionable for your team.

[Jamie Knight]

Hello, I don't see the impact ratings filter on my view. 

I'll keep an eye out! 

Also, I asked on a different article regarding the new release notes but never saw a response - will there be an option to subscribe to the release notes? 

[Jeffery Candiloro]

Why are you using 'AI' to generate the impact ratings? Surely, when planning/developing/scheduling a new feature/release there is some human assessment as to its likely impact on the apps and the clients? Can't this information be communicated rather than adding another layer of automated assessment that doesn't have the entire context?

[Charlie Misonne]

This overview is great. Thanks for sharing!

I do agree with @Jeffery Candiloro about the AI part. At least with the icon and "Uses AI. Verify results" it's transparent. I hope it stays.

[Julia Foden]

I don't see the Impact Ratings yet. It's a good idea to have another way of filtering the release notes, and useful info when doing a quick scan. But I fully agree with @Jeffery Candiloro  - why AI ? 

[Richard Scholtes]

@Jeffery Candiloro 

I'm normally front and center when it comes to critically second guessing decisions Atlassian is doing, but with the range and sheer mass of customers Atlassian has, a single-module focused developer or solution engineer will not do a better job here, as the impacts will differ heavily between customers per change. And how should AI not have the entire context? With all the data in Atlassian's own Atlassian platform it will a have ten times better overview of the context then any human ever could.

[Alex Chiu]

Hi @Jamie Knight 

The impact ratings were in the middle of a gradual rollout when you posted, but they should now be live for everyone. If you’re still not seeing them on your end, please let me know.

Regarding subscriptions: thank you for highlighting this! We know how important it is to easily stay on top of new release notes, and adding subscription/notification options is a top priority for our team. We’re actively exploring solutions and look forward to sharing more details soon.

[Alex Chiu]

Hi @Jeffery Candiloro 

That’s a fair question. The AI assessment doesn’t replace the judgment of the feature team responsible for the change. In fact, the team’s internal assessment is a key input into the Impact Rating.

AI helps us turn that information into a consistent classification across a large volume of release notes. We periodically compare the results with human judgment, review edge cases, and use feedback to improve the classification.

We also recognize that the rating doesn’t yet understand every customer’s unique organizational context. That’s why we clearly identify when AI is used and encourage customers to verify the results. Longer term, we’re exploring how Impact Ratings could become more relevant to each organization, so stay tuned!

[Alex Chiu]

Hi @Charlie Misonne 

Completely agree that transparency is critical, we plan to keep the AI label and verification guidance visible. 

To clarify how it works behind the scenes: the AI assessment doesn’t operate in a silo or override feature teams. In fact, the team delivering the change provides the foundational assessment, which acts as a primary input into the rating.

AI allows us to standardize that information into a consistent, reliable classification across a high volume of release notes. 

[Alex Chiu]

Hi @Julia Foden 

Impact Ratings were being released gradually when you posted this, but they should now be available to everyone. Please let me know if you still don’t see them.

On why we use AI: the assessment doesn’t replace the judgment of the team responsible for the change. Their internal assessment is a key input, while AI helps apply a consistent classification across a large volume of release notes. 

[Alex Chiu]

Hi @Richard Scholtes 

Scale and consistency are two of the main reasons we’re using AI here. It can bring together a lot of different sources of information to provide a consistent Impact Rating across a large volume of changes.

That said, the current rating doesn’t use an individual customer’s organizational data, so it may not reflect every customer’s unique setup or priorities. That’s why we clearly label the use of AI and encourage customers to verify the result. Longer term, we’re exploring how these signals could become more relevant to each organization.

[Jamie Knight]

Hi @Alex ChiuYep, it's now available. Thanks! Also, I look forward to a subscription option! Thanks so much.

[Jeffery Candiloro]

@Alex Chiu ,

Thanks for the added details.

Will there be an API for the release notes? That will allow us to pull some subset as needed? As part of our services to clients we provide details on upcoming changes/releases from Atlassian that they need to know about (based on our knowledge of their systems, configuration/tools used, environment, business priorities, etc). All this needs to be done manually at the moment. Having an API that allows us to pull and process the information rapidly and at scale would be amazing.

Jeffery