PaCT
The PaCT (Publics, Communications, Trust) Initiative is home to MISI’s research into the dynamics of informational trust in contemporary social life.
As traditional gatekeepers of information lose their footing and new ones emerge, PACT asks a foundational question: how do people come to trust the institutions, messengers, and systems that shape what they believe? Through this research, PACT builds an evidence base for understanding, and ultimately strengthening, the relationship between publics and the institutions meant to serve them.
At the center of this work is a close look at institutional trust: how people form judgments about government, journalism, and other public institutions, and what causes that trust to grow, erode, or shift over time. Rather than treating trust as a fixed trait, we examine it as something continually negotiated, as it is shaped by lived experience, community context, and the perceived reliability of the messengers themselves.
We also investigate the growing role of AI in these dynamics. As algorithmic systems increasingly mediate the flow of information, we examine how they reshape the pathways through which trust is built, tested, or broken.
Finally, we take seriously the narrative dimension of trust: the stories people tell themselves and others to justify what they believe. These stories are often as important as facts in determining what people accept as true. By studying the interplay of narrative, belief, and evidence, we offer insight into not just what people trust, but why.
Primary contact: Sarah Halford
Initiative Projects
Trust in Context
What do people think about the presence of data centers in their local communities? This study explores how residents in specific geographic locations learn to trust or distrust information about data centers. Using mixed-methods, we examine the formation of trust relations across three local information ecosystems in New England.
Co-PIs: Eric Gordon and Chris Wells
Co-Investigator: Sarah J. Halford
On the Porch
A hyperlocal large language model (LLM) for community for community knowledge and representation
Built with the Talbot Norfolk Triangle community in Boston’s Dorchester neighborhood, this chatbot is a resource for community safety and cohesion. The model, trained on structured 311 and 911 data, policy reports and documents, and hours of community meeting and conversation data, will provide information about what’s happening, what people are thinking about, and what people can do.
PIs: Eric Gordon and Chris LeDantec, in partnership with The Boston Project Ministries