Freiling’s publications identify the challenges that algorithms pose in understanding online information environments, such as social media, where researchers don’t know who sees what. “For example, if we do not know who saw a message about climate change, we cannot measure its effect on people’s perceptions of climate change,” she said. She is advancing understanding of how this lack of access to data hinders theory building. Science communication is critical to advancing AI responsibly, she said, because it helps to reach diverse audiences affected by AI and identify ethical issues.
