Establishing Guardrails on the AI Roller Coaster
MIT Center for Information Systems Research · Research Briefing · 17 September 2026
Based on interviews with 48 leaders from 28 organisations, the briefing looks at how organisations can move quickly with AI without causing harm. Tunic Pay is the featured case: a fintech that helps global banks protect customers from fraud, using task-specific predictive models for scoring and LLMs to build warnings from the specific counterparty facts surfaced during an investigation.
Read the briefing →