Why this assessment exists

GenAI vendor contracts are still maturing. Standard terms from OpenAI, Anthropic, Cohere, and Mistral routinely contain IP, data-use, and exit clauses that materially shift risk to the buyer. Customers who sign without a structured contract review pay 10–20% more and accept asymmetric data and IP terms.

This assessment maps your contract readiness against the patterns we've seen across 45+ GenAI vendor reviews since 2023. Built on direct experience with enterprise rollouts of OpenAI, Anthropic, Microsoft Copilot, and Google Gemini.

Your progress 0% complete
Question 1 of 8

Are IP / model-output ownership terms reviewed and acceptable?

Standard GenAI terms vary on whether output is licensed back to the customer or assigned. Material IP risk if mis-specified.

Question 2 of 8

Are data-use clauses (training, retention, sub-processor) explicit?

Default terms often allow training on your data or extended retention. Explicit opt-out and retention limits are essential.

Question 3 of 8

Has training opt-out / model fine-tuning posture been negotiated?

Most enterprise contracts can negotiate explicit no-training defaults. Fine-tuning rights and model versioning matter for stability.

Question 4 of 8

Are rate cards (per-token, per-seat, per-API call) benchmarked and locked?

GenAI rate cards drift; new model versions and pricing tiers appear quarterly. Locked rates with caps protect the case.

Question 5 of 8

Are SLA, availability, and incident-response terms commensurate with usage?

Production GenAI use needs SLA. Default consumer-grade terms are inadequate for business-critical workflows.

Question 6 of 8

Are indemnities (IP, copyright, data) symmetric and capped at appropriate levels?

Indemnity caps are typically asymmetric in vendor favour. Copyright indemnity is critical for content workflows.

Question 7 of 8

Are exit and portability terms documented (data export, prompt assets, fine-tuned models)?

Without explicit exit terms, switching cost is unbounded. Prompt libraries and fine-tuned models are at-risk assets.

Question 8 of 8

Is there clear governance ownership for the GenAI contract (legal + procurement + AI governance)?

GenAI contracts span legal, procurement, security, and AI governance. Single ownership prevents gaps.

0 of 8 answered

What happens next

When you click View your results, we'll ask for your name, work email, and company. We only accept corporate email addresses — no Gmail, Outlook.com, or other free providers — because this report is written for enterprise buyers and we use the domain to tailor the recommendations. Your email is never sold, shared, or used for anything other than delivering your report and (if you opt in) related GenAI research.

Once you submit, you'll be redirected to a personalised report showing your overall score, risk band, the specific findings for each question where you scored 2 or higher, and the three most important actions to take before you sit down with GenAI.

Prefer to walk through this with an expert?

Our GenAI practice will run the full diagnostic with you in a 2-hour working session.
Book a session →