After much discussion, the Reacting Consortium is rolling out its updated guidelines for use of LLMs/Generative AI by both Instructors and Authors. Those policies can be found on our AI Policies Page under Resources. In accordance with the "Requirements" section of the Instructor LLM Use Policy, we will be reaching out to authors in the coming weeks to secure their preferences regarding their games. What follows is a brief reflection by Reacting Consortium's Executive Director, Nick Proctor, on the process and principles that led to this updated set of policies.
By Nicholas Proctor
Generative AI, particularly large language models, presents important challenges for the entire Reacting project. Most of these challenges manifest in the classroom. Students find all sorts of ways to use AI. In my experience, most of those uses are bad, but your individual mileage may vary.
I am excited that the Reacting community continues to generate interesting ideas about student use of AI. We will certainly continue to provide opportunities to develop and discuss them, but deciding what your students may or may not do with AI is beyond the Reacting Consortium's authority. It's your classroom. You know it and your students best. Additionally, your institution may have its own AI policies. For these reasons, we decided fairly early on that the Consortium should not attempt to regulate AI use in individual classrooms.
Other questions, however, require collective action. In consultation with the Reacting Consortium Board (RCB) and Reacting Editorial Board (REB), and with advice from community members, especially John Giebfried and Ray Kimball, I prioritized developing two AI-use policies. The first governs the use of materials in the Reacting library. The second addresses the use of AI in creating new Reacting games.
I started experimenting with ChatGPT when it became publicly available and began sharing my thoughts more systematically in January 2025 when I presented “Exploring AI in Reacting Games: Challenges and Opportunities” at our Winter Institute. The following June, I presented “Reacting to AI: What if you want to go for it?” at the Summer Institute. Generally, I was enthusiastic about the possibilities, but my experiments with AI in my own classes revealed important shortcomings. Subsequent improvements in AI have alleviated some of these concerns, but they have generally not eliminated them.
As more people began using AI to design and draft Reacting games, it became clear that we needed guidelines. Personally, this was a great instance of the principle of Miles’ Law of bureaucracies: “Where you stand depends on where you sit.” As an individual, I was champing at the bit to explore the use of AI. As Executive Director, I was compelled to consider the Consortium’s obligation to defend our authors’ intellectual property.
Serious work on both policies began in the spring of 2026. I led efforts to develop the library policy, focusing on the use of LLMs to manipulate and adapt downloadable materials that have been written by our authors and reviewed by the REB. I coordinated with Kelly McFall and Emily Gray, the chair and co-chair of the REB, who oversaw the development of the policy governing the creation of new games. Developing these policies involved multiple drafts, online listening sessions, extensive email threads, and occasional face-to-face meetings to hash things out.
The latter were essential to the process. Long, wide-ranging conversations with RCB member Trey Alsup at Knutepunkt, Kelly and Emily at the 2026 GDC, and later with Kelly and RCB chair emeritus Tony Crider over shrimp and grits in Charlotte, North Carolina, shaped the documents and fed my soul. The deep skepticism about LLMs expressed by members of my family, and the “burn it with fire” position of Reacting Consortium Administrative Director Jenn Worth, also doubtless influenced my thinking.
Throughout this process, I was repeatedly impressed by the thoughtfulness, effort, and even-handedness with which dozens of community members engaged in these discussions. Disagreements could be sharp, and positions on AI use were sometimes antagonistic, but everyone agreed on our shared objective: helping students to learn.
Kelly brought the draft policy on new-game creation to the REB for final review and approval. He reported the results to the quarterly meeting of the RCB in October. Given the REB's editorial independence, this was an informational briefing.
I brought my final draft to the same meeting. RCB chair Ben Alpers and I agreed that adopting the policy fell within the Executive Director’s authority. Nevertheless, we thought it important to gauge the RCB’s support, particularly because the board represents the diverse interests of the Reacting community. My proposal included some possible adjustments, which the RCB discussed before voting to support the document. Having passed muster with them, I hope it meets your approval.
As everyone recognizes, AI is changing rapidly. Consequently, we built a one-year review requirement into both policies. They must be reviewed, and potentially altered, after a year has passed. If these policies delight or offend you, consider involving yourself in this ongoing process.
I mentioned lots of Reacting “insiders” in the narrative above, but I’m also struck by how many relatively new members of the Reacting community this process involved. They emailed me, spoke with me at conferences, and shared their ideas in online meetings. Reacting is continually strengthened and refreshed by new voices. I encourage you to add your own. It may sometimes sound like a cacophony, but to me it is a chorus.
