Another Hugo Awards-related controversy this year? Fans of sci-fi and related media may be thinking this after hearing today that three leaders of the Seattle 2025 Worldcon – the next iteration of the convention, where the Hugos will be presented annually – have resigned. The awards themselves, which are voted on by the World Science Fiction Society, seem to be out of the picture this year.
The award-winning World Science Fiction Society (WSFS) has voted on the winners of this year’s awards. The three announced their resignations from their respective roles in a post on Bluesky (19459013) that was co-signed by Hugo Administrator Nicholas Whyte and deputy Hugo Administrator Esther MacCallum Stewart. Cassidy is the World Science Fiction Society’s division head. The statement stated that “no LLMs or generative artificial intelligence have ever been used in the Hugo Awards Process at any stage.” This might be enough to turn the heads of those who are a) interested but b) unaware of the latest controversy.
But many people in the community know what’s going on. You can also get up to date by visiting the blog File770 as well as Seattle Worldcon 2025, which shared on April 30 a post that clarified the role AI played in this upcoming event.
“We’ve received questions about Seattle’s use AI tools in our vetting of program participants,” Seattle Worldcon chair Kathy Bond wrote. “In the interests of transparency, we’ll explain how we use a Large Language Model. We understand that many members of our community are concerned and have strong opinions about the use of LLMs. Please be assured that the LLM script used did not contain any data other than the name of a panelist. Repeat: no data, other than the name of a panelist, was entered into the LLM script. The sole purpose of the LLM was for streamlining the online search process for program participant vetting. Instead of being accepted uncritically by our team, the outputs of the LLM were carefully analyzed for accuracy.
Bond continues to explain in his post that “using this procedure saved literally hundreds hours of volunteer staff’s time, and we think it resulted to more accurate vetting once the step of verifying any purportedly negative results.” We have not used an LLM for any other aspect of the program or convention.
This last line was also highlighted in the resignation post today from Whyte and MacCallum Stewart. It was to ensure that Seattle Worldcon 2025 could more efficiently compile panels to offer to convention attendees.
Bond posted an additional message to the convention’s website on May 2 reiterating her regret about using ChatGPT for program vetting. “I regret also releasing a public statement that didn’t address the concerns of our communities,” she shared. “My initial statement about the use of AI tools for program vetting was incomplete and flawed, and I missed the most important points.” I am sorry for the harm that was caused.
But, as File770 pointed out the damage had already been done. The use of ChatGPT at Worldcon in any capacity created a social media furor. At least one Hugo nominee, Yoon Ha Lee’s Moonstorm ( ) was also prompted to withdraw their book from consideration. In a post on Blueskyfrom May 1, the author referred to the Worldcon blog entry of April 30, and stated that he was withdrawing his title from consideration.
In a post todayin response to File 770 announcing resignations, the writer wrote “All respect to them and I’m thankful for their work, Sorry [things] happened.”
Seattle Worldcon takes place August 13-17, while the Hugo Awards are handed out on August 16.
Want to know more about io9? Find out when the latest Marvel, Star Wars and Star Trek films will be released, what’s coming up for the DC Universe in film and television, and all you need to know about Doctor Who’s future.
