Guidelines for the Responsible Use of Generative AI in the Experience Cloud Community

[kautuk_sahni]

The Experience Cloud Community values genuine and high-quality contributions from its members. With the rise of Generative AI technologies, we acknowledge the opportunities and challenges they present. This document outlines our guidelines to ensure the responsible use of AI-generated content while preserving the integrity and trustworthiness of Community discussions.

1. Responsible Use of Generative AI - Generative AI is meant to enhance, not replace, human input. Users should ensure AI-generated content is accurate, relevant, and adds value to discussions. It is crucial to verify the accuracy of AI-generated responses to prevent misleading information.
2. Transparency and Disclosure - Any post or comment containing AI-generated content must include a disclaimer stating its AI origin. This transparency helps members understand the context of the content. Ensure the AI disclaimer is clear and prominently displayed within the post or comment to maintain transparency. Suggested disclaimer: "This response is inspired by/sourced from Generative AI."
3. Quality and Accuracy - AI-generated content must uphold the same standards as human-generated content, in that it is informative, accurate, relevant, and helpful. Before posting AI-generated content, users should verify its accuracy by cross-referencing with reliable sources. This verification helps prevent the spreading of misleading information. If sources are provided in the AI-generated output, they should also be reviewed and/or added to the published response.
4. Reporting and Enforcement Community Managers (CMs) will monitor AI use to ensure compliance:
   • Reporting Issues: Users can report non-compliant posts by expanding the content’s options menu and clicking on “Report”. Alternatively, they can message a CM directly. Reasons to Report: Lack of AI disclosure, inaccurate or misleading information, or lack of verification.
   • Review Process: CMs will review the flagged content to determine if it violates the guidelines. If necessary, the post will be flagged for revision or removed entirely. Further action may be taken against repeat offenders
5. Continuous Improvement We value any feedback you may have to refine and enhance these guidelines. Updates will address emerging issues and technological advancements.

Disclaimer: The image used in the banner is generated by Adobe Firefly.

[gkalyan]

AI is a great tool to enhance our experience and reduce our effort. Relying entirely on AI/GEN AI in its current form is not ideal. It is always useful to validate and fact-check the responses.

Empathizing with the user's context might help. Always check before posting. Is this information useful for our own consumption/resolution of the issue if we were in the same situation?

Finally, always remember this is a community to share knowledge and uplevel each other.

[stephentmerkle]

With things moving more and more to Agentic, it becomes increasingly tougher to validate and fact-check results returned by agents (or multiple agents behind the scenes) that are iteratively and autonomously building upon each other to generate outcomes and outputs in milliseconds.

One good thing is that agents prompt you at multiple steps in the process to get human feedback, but it doesn't often provide links to sources upon which its outputs are based on like Gen AI does. This is generally because the agents are already operating in a walled garden and basing all activities on a user-pre-approved knowledge base.