The Automotive Commerce Protocol (ACP) standardizes the data exchange between automotive dealers, original equipment manufacturers (OEMs), and the artificial intelligence systems operating between them. ACP defines canonical object types for leads, vehicles, deals, customers, and service events, plus the rules for how those objects move across system boundaries without silent loss. It is patent pending and open source at ACPspec.org.
Anyone who has worked in automotive knows this problem. The dealer side runs on a DMS. The OEM side runs on its own portal. The CRM is in the middle. AI vendors are now plugging into all three. Nothing agrees on what a lead looks like. Nothing agrees on what a vehicle inquiry should contain. The data layer is a battlefield.
What Problem Is ACP Solving?
ACP solves the broken exchange between dealers, OEMs, and the AI systems sitting in the middle. Today every integration is bespoke. A dealer connects to OEM A one way. The same dealer connects to OEM B a different way. AI vendors connect to both with their own translation layers. Every translation drops information.
The result is that AI-generated data degrades at every system hop. By the time a rich AI-tagged lead reaches the BDC agent, half the value is gone. Nothing crashed. Nothing alerted. The data just got worse. That is structured data loss in action.
What Object Types Does ACP Define?
ACP defines the canonical shape of the objects that actually move in automotive workflows. Each one has required fields, optional fields, validation rules, and a version.
- Lead. The structured representation of a customer inquiry. Includes contact information, vehicle interest, intent signals, and AI-generated metadata such as intent score and recommended next action.
- Vehicle. The structured representation of inventory. Includes VIN, trim, options, photos, and merchandising metadata.
- Deal. The structured representation of a transaction in progress. Includes vehicle, customer, financing terms, and AI-generated metadata such as deal status confidence.
- Customer. The structured representation of a known buyer. Includes history, preferences, and AI-generated context.
- Service Event. The structured representation of a service appointment or repair order. Includes vehicle, customer, scope of work, and AI-generated recommendations.
Every object has a defined shape that any ACP-compliant system can validate. That is the whole point.
How Does ACP Differ From ADF and STAR?
ADF and STAR are the two existing automotive data standards. ADF handles lead exchange. STAR handles service data. Both were built for a pre-AI era. Both move data through email parsing or point-to-point APIs that were never designed to carry AI-generated metadata.
ACP picks up where they left off. It can wrap ADF and STAR payloads where backward compatibility matters. It adds the AI-specific fields, the versioned envelope, and the fidelity score on top. Old standards keep working. New AI-aware standards stop the silent loss.
How Does ACP Use CORD?
ACP rides on top of CORD. ACP defines the contents of the object. CORD defines the envelope around it.
A complete ACP Lead is an ACP-shaped object wrapped in a CORD envelope. The envelope carries the schema version, the fidelity score, and the audit trail. Every system that touches the object can read the envelope to know what it is supposed to look like and can update the fidelity score if it had to coerce or drop anything.
Together they make AI-generated automotive data accountable across system boundaries.
Who Implements ACP First?
Adoption happens in three waves. Wave one is the AI vendors who already know the data layer is the problem. They implement first because they feel the pain hardest. Wave two is the platform layer. CRMs and DMSes that want to stop losing data implement next. Wave three is the OEM side, which is slower because OEM IT moves slower.
My own products implement ACP today. AutomotiveAI, DealersGear, and Drivevex all speak ACP internally. The reference implementations are open.
What Does This Mean for Dealers?
For dealers, the practical implication is that you should be asking your vendors whether they implement ACP. The vendors that say yes are committed to not losing your data at the integration layer. The vendors that say no are committed to keeping you locked into their proprietary shape.
This is the data sovereignty question for automotive. The standard is open. Vendor support tells you who is serious about the next decade.
Where to Read More
The full specification is at ACPspec.org. The companion universal standard is at CORDspec.org. The broader operating context is in my pillar piece on both standards.
Final Thought
The automotive industry has spent thirty years patching its data layer with bespoke integrations. AI is going to expose every gap. ACP is the standard that closes them. The dealers and OEMs that adopt early will run cleaner operations and have better AI outcomes. The ones that wait will spend the next decade rebuilding what they could have built right the first time.
Want the full system? The AI Operator Playbook ($499) is the operating manual I use to deploy AI across real companies. Structure, governance, and scale documented end to end.
Structure before tools. Strategy before speed. Systems before scale.
Ready to operationalize AI inside your business?
The AI Operator Framework is the system I use to design, govern, and scale AI for real companies. Start with the AI Operator Playbook ($499) or apply to work with me directly.
