New Study Shows for the First Time that AI Co-workers Significantly Enhance Forecasting Accuracy in Human Team Meetings

Research shows that "Proactive Scouting Agents" integrated into real-time group discussions resulted in an 18% improvement in team forecasting precision.

ARLINGTON, VA, UNITED STATES, September 17, 2026 /EINPresswire.com/ -- An important new study demonstrates for the first time that adding proactive conversational AI agents to real-time human team meetings can significantly enhance group forecasting accuracy. The findings are detailed in a newly released pre-print paper titled "Evaluating the Participation of Proactive Scouting Agents in Real-time Human Team Meetings".

Conducted by researchers from Unanimous AI and Carnegie Mellon University, the study explores the integration of AI co-workers into real-time forecasting discussions among small human teams. Unlike previous reactive AI tools that must wait for a user to ask a question before providing answers, these proactive “contributor agents” — also known as (Co)agents — act as autonomous members of the group meeting.

These agents are uniquely designed to monitor group conversations in real-time, independently identify critical knowledge gaps in the deliberation, autonomously search the open internet for appropriate information, and introduce this acquired knowledge into the discussion using natural, groupwise human etiquette. Importantly, (Co)agents do not replace human stakeholders or offer competing opinions – they are designed to support the deliberation of the human teams by finding and contributing meaningful, missing information, ensuring teams are well-informed during group discussions.

To test the impact of these proactive agents on group forecasting, researchers tasked numerous small human teams of four or five members with collaboratively forecasting the winner of Major League Baseball (MLB) games over a multi-week period. When assisted by a proactive scouting agent, team performance increased significantly.

Key findings from the study include:
• Raw predictive accuracy for the human teams increased from 60% without AI assistance to an impressive 76% when assisted by the (Co)agents.
• The inclusion of proactive scouting agents significantly reduced the groups' mean-absolute forecasting error by 18% (p=0.006).
• The proactive agents autonomously managed their participation effectively, accounting for 18% of the total conversation without disrupting human interaction.
• 86% of human participants (N=116) reported that the scouting agents were genuinely helpful to the group's forecasting efforts (p<0.001).
• 90% of survey respondents indicated that they believe enterprise business teams will want to incorporate these (Co)agents into future forecasting processes (p<0.001).

Overall, this study demonstrates that as long as human participants possess a baseline level forecasting skill, proactive agentic assistants can successfully amplify team decision-making capabilities by acting as a natural participant in the discussion. By autonomously mitigating information barriers and filling knowledge gaps in real-time, these agents serve as supportive and informative digital co-workers that do not replace human teams but ensure the human members are well-informed as they deliberate.

“The hardest part of building proactive agents for use in team meetings is emulating human group etiquette,” said researcher Hans Schumann. “For this reason, we were pleased to see that human teams not only made smarter predictions when (Co)agents participated but also appreciated their contributions, with over 80% of participants finding the agents’ comments to be timely, relevant, and helpful to group deliberation.”

This marks a significant milestone for human-AI collaboration, suggesting that the introduction of proactive agentic co-workers in real-time human meetings in platforms such as Microsoft Teams, Google Meet, Slack, or Zoom promises to usher in a new era of smarter, more deeply informed team decision-making.

About Unanimous AI
Unanimous AI develops innovative collaboration technologies that scale communication and amplify team intelligence. The company is the creator of Swarm AI^®, Hyperchat AI™, Hypervideo™, (Co)agents™, and Hyperlingual™ technologies and is the developer of the award-winning Swarm^® and Thinkscape^® platforms. For more information on Unanimous AI or its technology and platforms, visit Unanimous.ai.

Company Press Contact:
Sofia Spieler
Sofia@unanimous.ai

Sofia Spieler
Unanimous AI
Sofia@unanimous.ai