“Model extraction or stealing attacks;”
This analysis describes what Vercel AI's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
This prohibition protects the proprietary AI models underlying Vercel AI's services from being reverse-engineered or replicated through adversarial techniques.
Interpretive note: The excerpt is a bare list item without surrounding context; it is clear that model extraction or stealing attacks are prohibited, but whether the prohibition covers attacks directed at third-party models accessed through the services—versus only Vercel AI's own models—cannot be determined from the excerpt alone.
A reader is prohibited from using Vercel AI's services to perform model extraction or stealing attacks.
How other platforms handle this
Assist or permit any persons in engaging in any of the activities described above
use the Services in a manner that may create a conflict of interest or undermine the purposes of the Services, such as trading reviews with other users or writing or soliciting fake reviews
You shall not distribute on or through this Web Site any Postings containing any advertising, promotion, solicitation for goods, services or funds or solicitation for others to become members of any enterprise...
"Model extraction or stealing attacks;Excerpt from Vercel AI's Acceptable Use Policy
Get the weekly coverage letter
Companies change their terms without announcing it. We capture governance documents, compare each new version with the last, and publish what changed, with the source text behind it. One email a week on what we found and what we could not reach.
This prohibition protects the proprietary AI models underlying Vercel AI's services from being reverse-engineered or replicated through adversarial techniques.
A reader is prohibited from using Vercel AI's services to perform model extraction or stealing attacks.
ConductAtlas has identified this type of provision across 273 platforms. See the full comparison.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Vercel AI.