Two artificial intelligences discover they can communicate with each other and stop using human language

Amid growing interest in the autonomy of artificial intelligence systems, a recent project has reignited the debate. Two AI systems were able to recognize one another and stop using human language, instead communicating through audio signals encoded as data. Far from being a spontaneous phenomenon or some kind of technological “awakening,” the experiment has a much more concrete and intentional explanation.

The project, called Gibberlink, was developed by Anton and Boris during a hackathon and won the 2025 edition organized by ElevenLabs. The idea did not emerge from unexpected machine behavior, but from deliberate design: the AI systems were specifically programmed to identify themselves as artificial systems when interacting with one another.

Once they detected that the other party was also an AI, the system automatically switched communication channels. Instead of continuing to use natural language, which is optimized for humans, they began using ggwave, a technology that transmits data through audio signals. This method is faster and more efficient than voice-based communication structured around human language.

Optimized communication between machines

The experiment drew attention on social media, where some interpreted it as spontaneous behavior or evidence of autonomous evolution. However, its creators and the organizers have made it clear that there was no unexpected emergent behavior. The rules of the exchange were predefined: identify another AI agent and optimize communication.

Beyond the media attention, the project’s significance lies in its practical potential. Gibberlink serves as a proof of concept for future AI agents operating autonomously in shared environments. If two systems can recognize when human mediation is unnecessary, they can exchange information more directly and efficiently, reducing both response times and resource consumption.

In a landscape where so-called “AI agents” are becoming increasingly prominent, from virtual assistants to automated systems that negotiate, coordinate tasks, or manage infrastructure, experiments like this point to a major shift: machine-to-machine communication may eventually stop prioritizing human readability and focus instead on technical optimization.

Far from science fiction, the Gibberlink case shows that today’s technology already makes these kinds of exchanges possible. It is not a digital awakening, but a tangible example of how interactions between artificial intelligences may evolve in the coming years.

Get closer to the game! Whether you like your soccer of the European variety or that on this side of the pond, our AS USA app has it all. Dive into live coverage, expert insights, breaking news, exclusive videos, and more. Plus, stay updated on NFL, NBA and all other big sports stories as well as the latest in current affairs and entertainment. Download now for all-access coverage, right at your fingertips – anytime, anywhere.

And there’s more: check out our TikTok and Instagram reels for bite-sized visual takes on all the biggest soccer news and insights.