What questions should I ask an AI vendor before letting them into our systems?

The sales conversation with an AI vendor is designed to get you excited about capabilities. The questions that protect you are about control, portability, and what happens when things go wrong. These are the questions the sales team hopes you do not ask.

Data residency and access: where does your data physically live? Who at the vendor can access it? Can they access it without your knowledge? In what jurisdictions? This matters for GDPR, for industry-specific regulations, and for basic security hygiene.

Export and portability: can you export all your data, all fine-tuned model weights, and all derived analytics in open formats at any time? Not upon request with a 90-day processing window. At any time. If the answer is anything other than yes, your data is trapped the moment you upload it.

Model training rights: read the fine print. Many vendor agreements include clauses granting the vendor the right to use your data to improve their models. Your proprietary data is then training a model your competitors also use. Strike that clause. If they will not strike it, find a vendor who will.

SynthesisArc helps clients evaluate AI vendors during the INSIGHTS engagement. We review contracts, assess data governance practices, and recommend architecture that preserves your sovereignty regardless of which vendors you choose.