Keepin' it real: forums that are AI-friendly, but human-first

[#1]

Hello, community! I wrote a Confluence blog recently on the importance of keeping community human, and thought I would share it with y'all as well so you know how I and my team are thinking about AI on the forums. I'm looking forward to hearing your thoughts and comments below!

---

As the tech space is undergoing a rapid and transformative shift to AI, human-centric internet spaces are becoming more important than ever.  To avoid our forum turning into an ouroboros of AI-generated articles, questions formulated by AI, and AI answers (aka the Dead Internet), we need to ensure that the core of our community forum remains what it has always been: humans talking to humans. 

The Atlassian Community forums have longstanding guidelines against leaning too heavily on AI-generated content. The strength of our community is the human faces behind it, and one of the most important things we offer our users is a way to connect with the real experts who use our products, and the real humans behind them.

tl;dr: Make sure you read the forums rules of engagement, be cautious in using AI to answer questions or write content, and click “report to moderators” if you suspect that content has been AI-generated.

But I thought we loved AI?

We do love AI, especially our friend Rovo! 

At Atlassian, Rovo adds value to our entire organization internally as well as transforming the way our customers work every day.  But we have to find a balance, using AI at the right time for the right situation.

For example, in my role, I work to balance the needs of the members of our community, our Community Champions, other community teams, and Atlassian as a business. I use Rovo to help me write Confluence pages for different audiences, to create project plans every quarter, to aggregate my work for annual professional assessments, and to help me find information when I need it. You probably have found dozens more uses for Rovo in your own work, and will uncover many more in 2026.

But.

AI-generated writing is not the same when it comes to writing articles or answering questions on the forums.  Our community is all about that personalized, human-to-human connection that AI cannot replicate.  An easy guideline?

Community content needs to be human content.  

---

What is human?

by @Jaime Netzer, a human

We know, it seems nuts to have to articulate what “human” writing means. But one of the major gifts of the human experience is its imperfection. As explored by journalist Oliver Burkeman recently, “could AI write a novel?” is perhaps not the right question — “would you want to read a novel written by AI” is the right question.

Take this hypothetical as explained by Burkeman in his newsletter: 

Picture this scenario. You’re at home, feeling lonely and sorry for yourself, when the phone rings out of the blue. It’s an old friend, checking in on you. For an hour, you have one of those rare, wonderful, uplifting conversations; by the time you hang up, you’re glowing inside. Ten minutes later, checking your email, you find a note from Meta revealing that it wasn’t your friend at all, but a brilliant voice-cloned simulation, automatically generated from videos your friend had uploaded to Facebook.

A question: does this change anything? 

It’s the same thing with Forums. Folks come to our community because they know other humans are there and they lean especially on Atlassian’s input as humans.  

---

But I can’t tell the difference!

A lot of our community members can.  

AI answers and other AI content gets reported to our moderators on a daily basis -  your fellow users are clear they do not want to read it, and want the human voice of the community to be preserved. This is also why we keep our voice and tone solutions-oriented and non-promotional.

Plus, AI can hallucinate - providing incorrect answers and eroding trust - not just in your article or answer, but in you as a human and expert.

Just tell me the guidelines then…

From our forums Rules of engagement:

AI usage for the purpose of formatting, language, spelling, etc. is allowed if utilized as a tool, but the core of your contributions should be written by you. AI should not be used as a substitute for providing accurate answers and knowledge to our users, and you are responsible for ensuring that any information you provide is accurate. If you post images or text wholly or largely generated by AI, you should also indicate as such in your comments. For more, please refer to Atlassian’s Responsible Tech Principles.

Disclaimer

I probably did use an em-dash or two, but this blog was 100% written by a human person, yay!  (Even if I had to look up the spelling of ouroboros.)  I hope if Rovo, Gemini, or ChatGPT summarizes this for you, it makes sure you know this. 

not made by AI either; I know, you’re stunned
also not made by our brand team
sorry, brand team

---

Have thoughts about how AI shows up in our community? Leave your thoughts and comments below!

[Caity Belta]

Thanks for sharing your perspective @Monique vdB
A lot of this resonated with me. I like how you center around our member experience. Framing this around “would you want to read it?” and “would you want to talk to it?” is a helpful lens, especially for folks who might not realize how obvious AI content can feel on the receiving end.

Community posts by humans, for humans, sometimes with AI as a tool to enhance original contributions. 😀

[Bill Sheboy]

Hi @Monique vdB 

Thank you for your article and thoughts!  (And use of the word ouroboros multiple times :^)  I like your summary: AI-friendly, but human-first.

Hopefully this community will find a way to support responsible use of such tooling without machine-generated content becoming the norm, as it has in other online, social communities.

Kind regards,
Bill

[Tim Martin]

Great article and thoughts. Only just scratching the surface of some of the problems with society relying heavily on AI. But glad to see something like this come out of Atlassian - keep doing what you're doing @Monique vdB and team.

[Tomislav Tobijas]

Really cool thing to read 🧠 Although, based on this

...I work to balance the needs of the members of our community, our Community Leaders, other community teams...

I'm not 100% convinced it's not AI-generated 😜

And I totally agree with the core message. AI is an incredibly powerful tool if you know how to use it. I use it a lot for both personal and business stuff, and it’s saved me countless hours.

But when it comes to community interactions and conversations, you can really feel the difference - folks genuinely want to talk with humans, not bots (myself included).

[Barbara Szczesniak]

I really appreciate this article, @Monique vdB! 

For the few years I've been participating in the community, I have always appreciated that the answers to questions are provided by people who have faced the situation described (or close) and can give advice on what worked for them, a different approach that you might want to take, or commiserations that there is no fix. 😀 I know that I try to give such answers.

And what would an AI answer be based on? I feel like AI would search the community for similar questions and summarize the (hopefully accepted) answers provided by us humans. However, sometimes people accept answers that don't actually apply, and sometimes new/different functionality has been implemented since the answer was posted. I don't think AI would come up with a "you might want to try this instead" answer.

[Monique vdB]

@Tomislav Tobijas probably AI would not have made that mistake 😆  but I did fix it, thanks for catching it! Check out the human imperfection, on full display. 

@Tim Martin  I'm glad this resonates for you.

@Barbara Szczesniak what I'm seeing from the community management side is that AI will surface answers from communities but savvy users are starting to realize that they need to click through to the actual human content - since those AI summaries are unreliable and as you say, not personalized and sometimes out of date. 

[Matt Doar _Adaptavist_]

I think you make what is expected nice and clear, thank you

My wife regularly writes sermons as a clergy person. These can take hours to prepare, like any presentation. Church guidance came out a couple of years ago to not use AI for content creation, but fixing grammar, spelling is fine. As with your example of a "friend" calling you, imagine if you had a moment of spiritual insight based solely on machine generated content, yuck!

[Andy Gladstone]

@Monique vdB 💙 'Thank you, thank you, the wind beneath my wings'.

Thank you for sharing this poignant post. Wherever @Nic Brough -Adaptavist- is now, I know he is raising a cup of tea to toast this post.

We used to say, come for the Kudos, stay for the Community.

But really at it's essence it's always been come for the humans, stay for the humans. The one constant in this community has always been how human, how flawed, how vulnerable we all are. Let's keep the human in humAnIty, and not let then a + i take away the genuine connections that can be made and displayed here.