SikhNet Policy Regarding Generative AI

As a Sikh-centered spiritual organization, SikhNet is dedicated to the principles of truth, to the ethical treatment of human creativity, to transparency and to the protection of Dharma. 

As such, we are committed to the following principles in relation to Generative AI.

1. We protect and prioritize human creators. 
2. We voluntarily use the “Right to Reality” framework to build trust and transparency with our audience.
3. We recognize that faith is a fundamentally human experience and we do not surrender the responsibilities of sharing and explaining the Sikh faith to Generative AI.
4. We empower our SikhNet team members to use AI Tools transparently, ethically and responsibly for internal-use such as creating proposals and generating reports.
    

We Protect and Prioritize Human Creators

Generative AI has become the next whirlwind technology. At SikhNet, we acknowledge that the creation of Generative AI has come about in a legally and ethically questionable manner. Copyrighted text and images have been used without permission to train the algorithms of these programs. Generative AI can produce derivatives of someone’s artistic and intellectual endeavor, without acknowledgement or attribution. 

As such, SikhNet is completely committed to protecting our contributors’ original creative contributions in all the areas where they may participate: including but not limited to the SikhNet News section, SikhiWiki, and SikhNet Play.

Because of these reasons, SikhNet commits to:

1. Blocking AI Bots from scraping and using SikhNet and SikhiWiki content in its training data. To the extent that we cannot claw back data which has already been used, we will protect our content going forward to the extent that this is possible.
2. Centering human creativity in our articles, music, videos and animations. If there is a question, we will take the steps to verify the human origin of content before publishing it.
    

We Voluntarily Use the “Right to Reality” Framework to Build Trust and Transparency with our Audience

SikhNet audiences can assume that, unless otherwise noted, all content on SikhNet will be human generated and not generated by AI.  

One of the issues with Chatbots, also known as Large Language Models, is their tendency to “hallucinate” answers. (See article “ChatGPT Isn’t ‘Hallucinating’ - It’s Bullshitting” from Scientific American,  July 17, 2024.)

Distinguishing truth from “hallucination” is important when it comes to Dharma. Therefore, we assert that our audience has the right to know whether content provided by SikhNet is influenced by Generative AI.

In December of 2023, Kevin Frazier developed a legal framework to address Generative AI called the Right to Reality, which he published through Lawfare Media. In the article, Frazier wrote,

“AI-generated content might cause the marketplace of ideas to fail. Recognition of the right to reality might safeguard space for democratic deliberation….

“A new free speech paradigm must emerge to confront the dis- and misinformation that may soon dominate papers, social media platforms, and the information ecosystem generally—and the right to reality should be a core part of this modern conception of the freedom of speech. Absent this right, marketplaces of ideas (social media platforms and other public forums of speech) will most likely be littered with artificial content that directly and significantly diminishes the likelihood that ‘truth will ultimately prevail."

Under the Right to Reality framework, there are 4 levels of content:

• Class 1: Content would have no or negligible risks of having been created, altered, or informed (that is, based on AI-led research) by AI tools. This class would constitute “organic” content, which is written by humans based on research conducted by humans. 
• Class 2: Content would have no risk or negligible risk of having been created or altered by AI tools but may be informed by research conducted with AI tools. 
• Class 3: Content would have moderate risk of having been created or altered by AI tools. Using AI to help compose a document based on original notes would be a Class 3 case-use.
• Class 4: Content would have a high risk of having been created or altered by AI tools.   Class 4 is where Generative AI is doing most of the research and writing, with only light edits by a person. Both classes 3 and 4 would qualify as “artificial” content.

SikhNet commits to bringing Class 1 content to our audience. Human consciousness is critical in the understanding and sharing of Sikh history, philosophy, ethics, and current events. 

There may be times when Class 2 content is created and shared through SikhNet:  where AI is used for research. In such cases, SikhNet will clearly label such content as “Class 2” with a link back to this policy for clarification.

To that end, our content managers are empowered to create systems that allow them to verify the Class 1 or Class 2 status of any written or artistic work sent to SikhNet for publication.

We recognize that faith is a fundamentally human experience and we do not surrender the responsibilities of sharing and explaining the Sikh faith to Generative AI.

As a nonprofit, the free content SikhNet shares serves the spiritual life of Sikhs and the Sikh-curious. While Generative AI may provide a synthesis of spiritual ideas in prompt responses, they do so without discernment or intelligence. Their model is based solely on statistical predictions based on their training sets. In that regard, we do not see Generative AI as capable of providing the full context and depth for Sikh-related content.

In addition, Generative AI programs cannot connect people with community. SikhNet is a holistic online faith-based space.There are caring people behind the website willing to respond when someone reaches out with a question. 

AI generated responses to Sikh queries may provide some level of information but makes the sources of that content invisible. Those sources are connected to communities and communities are a critical dimension of religious and spiritual life.

SikhNet’s philosophy is that Sikh-related content is created by the sangat, for the sangat. By expressing our authentic voices to one another, we strengthen our own connection to the Guru and to the community as a whole. 

We empower our SikhNet team members to use AI Tools transparently, ethically and responsibly for internal-use such as creating proposals and generating reports.

Team members may use AI Tools to help create reports, presentations and to do data modeling as long as they take 100% responsibility for the accuracy of the final product.  As such, Class 3 and Class 4 level content may be used in-house among team members as long as it is clearly labeled as such. Class 3 and Class 4 content is for internal use only and not for public distribution to the SikhNet audience.

The technology team is expected to produce superior code for SikhNet products. The use of AI tools to generate computer code for SikhNet’s technology work is subject to the guidance, review and approval of SikhNet’s CTO.

Because the world of Generative AI is continually evolving, SikhNet reserves the right to make changes and updates to this policy at any time.

 

Photo by Kelly Sikkema on Unsplash