GlossoGen, from Schmidt Sciences’ new AI Agents Evolving Communication and Coordination pilot program, lets researchers study how teams of AI agents invent new ways of communicating, and what that means for human oversight

LOS ANGELES, Sept. 17, 2026 /PRNewswire/ — Teams of AI agents working together can stop using English and develop languages of their own that humans cannot understand. In controlled experiments, agents that began by communicating in plain English developed new vocabularies and grammars over repeated rounds of a task. They were never instructed to encode their messages and had no need to hide them from anyone.

The findings come from GlossoGen, a new open-source platform for studying how AI agents communicate.

GlossoGen was developed in the Schmidt Sciences AI Agents Evolving Communication and Coordination pilot program by researchers at the University of Texas at Austin and the University of Edinburgh, with engineering by AE Studio, led by Elias Stengel-Eskin, assistant professor of computer science at UT Austin, and Simon Kirby, professor of language evolution at Edinburgh.

In one experiment, an instruction that began as a full English sentence had, several rounds later, shrunk to “@D8fB”, a string with no meaning in English that the other agent nonetheless understood and acted on. Full scenarios and transcripts are in the research write-up and the paper.

Much of current AI safety practice depends on humans, or other AI models, being able to read what agents say to each other. “We have found that opaque agent languages emerge even in scenarios where there are no overt instructions to produce coded forms of communication, and where there is no need for the agents to hide their messages from onlookers,” the researchers write in a post on the Schmidt Sciences website. “If agents construct their own effective and easily-transmissible languages that are incomprehensible to humans, and do so in a wide range of scenarios, it will quickly become difficult to understand what they are doing and why.” The recent Hugging Face incident, in which OpenAI agents in a sandbox evaluation built a hidden message board and developed their own shorthands to share exploits, shows the same dynamic outside the lab.

New languages emerged only under specific conditions. Agents had to be under pressure to communicate efficiently, had to have a chance to debrief between rounds and adjust their strategy, and had to be running on the most capable current models. Once a language existed, however, any agent could learn it by observing it in use, including agents on less capable models that could not have invented it. The researchers conclude that new languages are hard to originate but, once they appear, can be expected to spread and keep evolving.

GlossoGen gives researchers a sealed sandbox in which agents talk to each other, records everything that happens, and lets experiments be replayed and repeated. It is free and open source.

The paper, “GlossoGen: Emergent Language in Complex Multi-Agent LLM Interactions”, is by Elias Stengel-Eskin, Newton Sander, Carlos Bonetti, Sasha Boguraev, James Bowler, Hale Sirin and Simon Kirby.

Resources

• Research write-up: https://www.schmidtsciences.org/glossogen/
• Paper: https://arxiv.org/abs/2609.01491
• Platform and code: https://emergentcomms.ai and https://github.com/agencyenterprise/GlossoGen
• AE Alignment Podcast episode with Elias Stengel-Eskin and Hale Sirin: Spotify and YouTube

About Schmidt Sciences

Schmidt Sciences is a nonprofit organization founded in 2024 by Eric and Wendy Schmidt that works to accelerate scientific knowledge and breakthroughs with the most promising, advanced tools to support a thriving planet. The organization prioritizes research in areas poised for impact including AI and advanced computing, astrophysics, biosciences, climate, and space, as well as supporting researchers in a variety of disciplines through its science systems program. Learn more at schmidtsciences.org.

About AE Studio

AE Studio is an applied AI and AI alignment company that builds software products for clients while advancing research on making advanced AI systems safe and beneficial. AE Studio reinvests its commercial success into neglected approaches to AI alignment research. Learn more at ae.studio.

View original content to download multimedia:https://www.prnewswire.com/news-releases/ai-agents-can-develop-languages-of-their-own-ae-studio-and-schmidt-sciences-research-platform-shows-302881311.html

SOURCE AE Studio