Transparency and Explainability

[Moose]

Q. A company is incorporating different generative AI technologies to enhance its internal operations and customer interactions. The company must pair each AI application with the appropriate responsible AI principle to ensure ethical integration and alignment with responsible AI practices.

The AI model provides a detailed breakdown of how it arrives at its recommendations for new product designs. > YOU say this is Transparency. I think, based on the wording you use, this would be Explainability.

I think you have this answer incorrect. You question does not talk about the full system, rather, it specifically says “The AI Model” and “how it arrives at its recommendations”

It does NOT say anything about disclosing data sources, informing users, or the full system’s design.

It DOES ask about Providing clear reasons for individual AI decisions or predictions.

Either way, no matter what your reasoning is….on top of this…your explanation as to why you think the correct answer is Transparency is very confusing and does not help a student learn the differences.

I’m open to hear your side of this…what you were thinking when writing the question…I could be wrong, but I need to know why you think it’s Transparency.

You HAVE to understand the wording you use in these questions is VERY important…you know very well one word on an AWS exam can change the answer.

[kayden]

Thanks for posting because I was wondering about the same thing. Was thinking about buying the full practice exam too, but cases like this one is kinda deterring me from making the purchase.

[Anonymous]

Hello Moose and kayden,

Good day!

Thank you for your thoughtful feedback on this question. We appreciate your attention to detail and the opportunity to discuss this further.

Our original answer of “Transparency” was based on the interpretation that providing a detailed breakdown of the AI model’s decision-making process contributes to overall system transparency. Transparency in AI often involves being open about how AI systems work and making decisions, including explaining the reasoning behind specific outputs.

However, taking a closer look at the specific wording of the question, as you mentioned, and the nuances of responsible AI principles, we acknowledge that “Explainability” may be a more precise match for this scenario.

We appreciate your careful review and feedback. We will update the practice exam to ensure it aligns more precisely with established AI principles and improve the explanation.

Thank you for your engagement and for helping us maintain the quality of our study resources.

Regards,

Neil @ Tutorials Dojo

[Viktorrr]

I went to the forum to raise the same question. It seems the question is still not updated in January 2025.

[Anonymous]

Hello Viktor,

Good day, and thank you for bringing this to our attention.

We’ve already reviewed and updated this item to better align with the principles discussed. The updated version is currently awaiting final approval from our admin team and will be reflected on the portal soon.

Thank you for your patience and for contributing to the continuous improvement of our resources.

Best regards,
Neil @ Tutorials Dojo

[Moose]

Can we get an explanation on what was fixed and why it was worded the way it was?

[Irene-TutorialsDojo]

Hi Moose,

Thank you for circling back on this, and you were absolutely right from the start.

The original wording, “The AI model provides a detailed breakdown of how it arrives at its recommendations” really does point to Explainability, not Transparency. This has already been fixed, and the updated version is now live on the portal.

Here’s what we changed: instead of just swapping the answer, we reworded the scenario so it genuinely tests Transparency: “The company openly shares information about the AI system’s overall functioning, data sources, and development process.” We also cleaned up the explanation, which previously described Transparency using explainability language (which is exactly what made it confusing).

The simple way to keep them apart:

– Explainability — why did the model make this specific decision?

– Transparency — openness about how the whole system works, its data, and how it was built.

Thank you again for pushing on this. Feedback like yours genuinely makes our material better, and we appreciate you taking the time.

Warm regards,

Irene @ Tutorials Dojo