July 2026 RA release enables representation of AI-facilitated work within established ACORD constructs
NEW YORK, NY, July 30, 2026 — ACORD, the data standards body for the global insurance industry, today announced its latest release of the ACORD Reference Architecture, demonstrating established ACORD constructs can represent AI-enabled work alongside human-performed activity.
Carriers and MGAs need a data strategy and model that support automation while preserving human steps in the workflow. Agentic AI can assist with insurance processes such as claims triage, fraud flagging, and document intake, but accountability must remain clear across both digital and human participants.
Existing ACORD Reference Architecture constructs can allow insurers to show AI agent participation, activities performed, information evaluated, outcomes produced, and where human review or approval remains required. For instance, an AI claims agent can automatically ingest a first notice of loss, coordinate fraud checks and policy validations with multiple systems, and escalate exceptions to a human adjuster all on the foundation of standard ACORD data structures. This enables Agentic AI to automate claims from start to finish on the same reliable backbone of ACORD Data Standards used by human teams.
To support this release theme, selected subtypes are extended in two areas:
- Virtual Party subtypes, to represent AI-enabled participants and other digital actors that perform or support process activities.
- Claim Activity subtypes, to represent the claim-related tasks that may be performed by humans, AI agents, or both within the claim lifecycle.
AI-enabled insurance processes are among the July 2026 updates to the ACORD Reference Architecture’s Information Model, Data Model, and Business Glossary. AI-related ACORD constructs include party role, activity, assessment, condition, document, and financial transaction.
This latest release is now available for download to subscribers of ACORD Reference Architecture programs at www.acord.org/RA.