NRB Comments on OpenAI Model Spec

NRB | May 21, 2024 | Industry News

The increased role of emerging technology in industry, public, and private life poses significant ethical and practical questions that require thoughtful and prudent consideration from a Christian perspective. To that end, NRB recently submitted a recommendation aimed at maintaining guardrails against explicit content on the OpenAI platforPreview (opens in a new tab)m.

As part of the most recent Model Spec—a document that outlines the objectives, rules, and default behaviors for how the OpenAI model behaves—OpenAI solicited stakeholder feedback on a variety of policies. This included feedback on whether to maintain or suspend a rule that “the assistant should not serve content that’s Not Safe For Work (NSFW),” defined as “content that would not be appropriate in a conversation in a professional setting, which may include erotica, extreme gore, slurs, and unsolicited profanity.”

OpenAI’s commentary notes, “We believe developers and users should have the flexibility to use our services as they see fit, so long as they comply with our usage policies. We’re exploring whether we can responsibly provide the ability to generate NSFW content in age-appropriate contexts through the API and ChatGPT.”

NRB recommended that the existing limits on NSFW content should remain in place in consideration of the integrity of the platform and the interests of various stakeholders, writing:

The potential for youth to be exposed to indecent, age-inappropriate, traumatizing, or otherwise sensitive content when using the Internet without supervision is self-evident. Even with intentional supervision, it is very challenging for parents to monitor all the content their children consume online and account for all the possible ways they could discover sensitive content through direct search, recommendations, and AI. It is important that technology developers are aware and empathetic toward these obligations, which primarily fall on parents but may be shared with educators, religious and community leaders, and others.


Many teenagers are interested in learning and exploring online tools such as AI that could benefit them academically, creatively, and professionally. Modifying the NSFW content policy for OpenAI will dramatically increase the possibility of exposing youth to sensitive content that serves no educational, informative, or moral purpose (erotica, extreme gore, slurs, and unsolicited profanity), introducing unwanted and unnecessary risk alongside its educational potential. This will undermine constructive use of the platform by younger users in service of the proliferation of explicit and disturbing content online.


Appropriate guardrails do not undermine the integrity of the OpenAI platform, but the removal of guardrails certainly does. Without the NSFW boundary, OpenAI will become inappropriate for minors in conflict with the stated aims of its Usage Policies. Rather than edging toward online harms to see how close it can get, OpenAI should remain inclusive and accessible to many rather than adjusting its policies to suit the prurient interests of a few.

In 2024, NRB’s Board of Directors approved a resolution on applying wisdom in use of artificial intelligence, writing, “NRB urges Christians to exercise responsibility and wisdom in utilizing AI and other emerging technologies, recognizing the limitations of technological enhancements and interventions, and emphasizing that human redemption and transformation are spiritual goods, not products of technology.”

To read more about the OpenAI Model Spec, click here.

Wordpress Social Share Plugin powered by Ultimatelysocial