Does anyone else not use any of the AI features in the platform?

I have deactivated all the AI stuff in our platform.

Very interesting question. I’m interested to see what other users say as well. Here’s a glimpse at what we currently have activated/partially activated in the platform.

I’m very intrigued by all the Harmony capabilities and by Docebo’s MCP, but the biggest barriers to adoption for both solutions for us are:

1. User Experience testing – Testing & validating the user experience to a point that we feel comfortable turning them on
2. Naming/branding – Aligning on naming/positioning of these capabilities. I don’t think we want to call it “Harmony”, so we’ve been exploring alternative names and trying to determine which keys in the Localization Tool we’d need to translate to call it something different, and ensure those names get used everywhere
3. AI Everywhere – Not sure if other organizations are experiencing this, but our company is struggling at the moment with what I’ll call “AI point solutions” popping up in all of our self-service tools. Our users have so many options to choose from when it comes to a natural-language processing interface where they can ask open-ended questions. This feels like the problem of non-federated search experiences on steroids. These chat experiences are meant to make it easier for users to find answers to their questions, but unfortunately the reality is instead their confusion about when to ask what question where. That’s part of the reason I’m so intrigued by the MCP and am currently testing it via Claude to see how good it is at information parsing & distillation

​@emily.mccarthy what are your top reasons for disabling everything?

Very interesting question. I’m interested to see what other users say as well. Here’s a glimpse at what we currently have activated/partially activated in the platform.

 

I’m very intrigued by all the Harmony capabilities and by Docebo’s MCP, but the biggest barriers to adoption for both solutions for us are:

1. User Experience testing – Testing & validating the user experience to a point that we feel comfortable turning them on
2. Naming/branding – Aligning on naming/positioning of these capabilities. I don’t think we want to call it “Harmony”, so we’ve been exploring alternative names and trying to determine which keys in the Localization Tool we’d need to translate to call it something different, and ensure those names get used everywhere
3. AI Everywhere – Not sure if other organizations are experiencing this, but our company is struggling at the moment with what I’ll call “AI point solutions” popping up in all of our self-service tools. Our users have so many options to choose from when it comes to a natural-language processing interface where they can ask open-ended questions. This feels like the problem of non-federated search experiences on steroids. These chat experiences are meant to make it easier for users to find answers to their questions, but unfortunately the reality is instead their confusion about when to ask what question where. That’s part of the reason I’m so intrigued by the MCP and am currently testing it via Claude to see how good it is at information parsing & distillation

​@emily.mccarthy what are your top reasons for disabling everything?

Very interesting question. I’m interested to see what other users say as well. Here’s a glimpse at what we currently have activated/partially activated in the platform.

 

I’m very intrigued by all the Harmony capabilities and by Docebo’s MCP, but the biggest barriers to adoption for both solutions for us are:

1. User Experience testing – Testing & validating the user experience to a point that we feel comfortable turning them on
2. Naming/branding – Aligning on naming/positioning of these capabilities. I don’t think we want to call it “Harmony”, so we’ve been exploring alternative names and trying to determine which keys in the Localization Tool we’d need to translate to call it something different, and ensure those names get used everywhere
3. AI Everywhere – Not sure if other organizations are experiencing this, but our company is struggling at the moment with what I’ll call “AI point solutions” popping up in all of our self-service tools. Our users have so many options to choose from when it comes to a natural-language processing interface where they can ask open-ended questions. This feels like the problem of non-federated search experiences on steroids. These chat experiences are meant to make it easier for users to find answers to their questions, but unfortunately the reality is instead their confusion about when to ask what question where. That’s part of the reason I’m so intrigued by the MCP and am currently testing it via Claude to see how good it is at information parsing & distillation

​@emily.mccarthy what are your top reasons for disabling everything?

For us, the top reasons for disabling these capabilities right now come down to governance and compliance.

1. If permission was every granted we certainly would undergo vivid testing for user experiences and outputs in a controlled environment, which for us would definitely be our sandbox. At this time, we can’t try this though as it still a risk for information being used that could potentially be transferred.
2. This is more so for you. Before coming back to this post, I just saw another post indicating that the ability to rename or rebrand Harmony is planned as a new feature being released on August 31st. Harmony's name is retiring. Yours takes its place. | Community
3. Our organization has very strict AI governance requirements. We currently limit AI usage to Copilot, as it aligns with our security, compliance, and confidentiality standards. Because of that, we do not introduce additional AI tools or chat experiences. There has been discussion about separate governance reviews for additional tools, but I have not been able to advocate strongly for this to occur at this time. Our Copilot integration is still in its’s first phases. 

Those are the primary reasons we're taking a more conservative approach today, though we're definitely keeping an eye on the roadmap and future enhancements.

Yeah basically not embracing any of the AI stuff where we are not being pushed into it just as of yet. To your point ​@emily.mccarthy - we are in a very similar groove with the small betas of AI tools disappearing around the org. v1 Governance has been developed and its first messaging about CoPilot being the tool of choice is getting resolved (personally, I have been challenged with the name CoPilot is all over the place and that this is the reality we have to help contribute and train into).