As part of its Yearlong Theme on Joint Human–AI Systems in an Agentic World, the Translational Data Analytics Institute (TDAI) hosted an engaging session exploring how artificial intelligence is reshaping the way we write—and how we might recognize its influence.

Dr. Mohit Iyyer of the University of Maryland offered a thoughtful, accessible look at the growing presence of AI-generated and AI-assisted text in everyday communication, creative work, and journalism. Rather than focusing solely on technical detection methods, Dr. Iyyer centered the broader challenge: as humans increasingly collaborate with language models, the line between human and machine authorship becomes harder to trace, raising questions about transparency, originality, and trust.

The event invited participants into that ambiguity through an interactive exercise where attendees examined short passages and tried to determine whether they were written by humans, AI systems, or a mix of both. The activity highlighted just how subtle—and sometimes surprising—the signatures of AI-assisted writing can be.

Throughout the talk, Dr. Iyyer emphasized that the goal is not simply to “catch” AI, but to better understand the complex ways people are beginning to co-write with advanced models. This shift, he noted, requires new frameworks for evaluating contribution, attribution, and responsibility in writing.

Hosted and facilitated by TDAI, the session brought together researchers, students, and interdisciplinary partners for an open, future-focused discussion that underscored why detecting and characterizing AI use is becoming an essential part of understanding modern communication.