Jerusalem, 29 January, 2026 (TPS-IL) — A new artificial intelligence system designed to track how cats interact may do more than improve animal welfare. And Israeli scientists say the same technology could one day be used to study social behavior in humans, after revealing hidden patterns in feline relationships that traditional human observation has missed.
One of the researchers behind the study explained to The Press Service of Israel that the project moves animal behavior research beyond traditional observation and into a data-driven era. Instead of relying on human interpretation alone, the team from Israel’s Haifa University used AI to track how cats physically relate to one another over time, revealing social patterns that are difficult or impossible for the human eye to detect consistently.
“In this study, we’ve shown that artificial intelligence can help us make the leap from animal psychology to animal sociology research, and it has practical implications in any field related to animal well-being, at home or at shelters,” Dr. Teddy Lazebnik from the Department of Information Systems at Haifa University and the Department of Computing at Jonkoping University, Sweden, told TPS-IL.
The study was recently published in the peer-reviewed Journal of Veterinary Behavior. At its core is a simple but powerful idea: measuring distance. Using AI-based computer vision, the researchers analyzed video recordings of adult cats interacting naturally in a controlled environment. The system continuously measured how close cats were to one another, how often that distance changed, and how sharply it shifted during different types of encounters.
In total, the team examined interactions among 53 adult cats, analyzing 186 distinct social encounters. The results revealed clear and consistent patterns. Female cat pairs tended to remain in the closest proximity to one another, while male pairs maintained the greatest distance. Mixed-gender pairs fell in between. Just as telling was how distance behaved over time. Friendly interactions showed relatively stable spacing, while tense or aggressive encounters were marked by rapid, frequent shifts between approach and withdrawal.
Beyond academic insight, the researchers say the technology could have immediate real-world uses. In animal shelters or multi-cat homes, social tension is often recognized only after aggression or stress-related behavior appears. An AI system that tracks subtle changes in distance and movement could flag unstable social dynamics early, allowing caregivers to intervene before conflicts erupt.
“Animal shelters can be very crowded,” Lazebnik said. “A tool like this can help caregivers understand the social dynamics between cats and decide where to place them.”
He added that the same approach could eventually be adapted beyond animals. By quantifying patterns of proximity and interaction, the technology could open the door to deeper research into human social behavior, offering a new, data-driven way to study how relationships form, shift, and break down over time.
“The research enables the expansion of this technology to humans and allows for deeper study of social interactions,” he said.
Lazebnik added that his team has already started follow-up research in dogs, and that data will soon be gathered on horses as well.
“It is a methodological leap from the individual animal to the group, giving us a much better understanding of their dynamic than before,” he said.
Related Topics
