US nonprofit CodePath and AI safety company Anthropic have announced the launch of a Knowledge Network designed to share live research findings on how AI tools affect student outcomes with decision-makers in education, workforce development and philanthropy before findings are made public, as reported by Edtech Innovation Hub.

The initiative was announced at the ASU GSV conference and is now recruiting founding members for a spring and summer launch. It is built around a 15-month observational study running across thousands of students at more than 1,000 colleges across the United States.

The research agenda is structured around two core questions. The first examines whether AI tools produce different effects across student subpopulations, including who benefits, who over-relies on AI, and what design features moderate those outcomes. The second investigates how the type of AI scaffolding predicts performance on unassisted assessments, and whether that relationship varies by prior experience or background.

Michael Ellison, Founder and CEO of CodePath, said the Knowledge Network is designed to close the gap between evidence and decision-making. "The idea is simple. Reliable findings shared with decision-makers in real time, while decisions are still being made," he said.

Madison McCormick, Strategic Initiatives Lead at CodePath, described the model as open source for evidence. "CodePath and Anthropic will be running a continuous research agenda inside real classrooms at scale," she said. "Organizations involved will get findings before they are public, and feed what they are seeing back into the next cycle."

The initiative carries an explicit equity focus, consistent with CodePath's mission of serving first-generation and low-income students.

Read the full details of the Knowledge Network and its research agenda in the full story.