AI vendors and your data

146 vendors · 427 answers

Does it train on API and business data?

Whether prompts and outputs sent through the API or a business plan are used to train models. 97 of 146 vendors answer it on the pages watched here, quoted below.

15 of 97 answers
Depends
Atlassian
Only de-identified metadata(Depends)
“Atlassian may use metadata to fine-tune open-source models that operate strictly within Atlassian's infrastructure, solely to improve the quality of responses and experiences we deliver to customers...”

AI Trust On the page Oct 26, 2026

Depends
Baz
Only de-identified, aggregated data(Depends)
“This data may be used for both internal and external purposes, such as analyzing, enhancing, testing, training and improve Baz Services (including through the use of artificial intelligence) and for research purposes...”

Baz Privacy Policy On the page Sep 25, 2026

Depends
Cohere
Opt out in settings(Depends)
“You can opt out from your prompts and generations being used to train Cohere models in your dashboard settings at any time.”

Enterprise Data Commitments On the page Oct 26, 2026

Depends
Dropbox
Feature models only, not generative(Depends)
“These models may be trained on your documents and metadata, and power features within Dropbox such as improved search relevance...”

Dropbox Privacy Policy FAQ On the page Nov 9, 2026

Depends
Figma
Off for Org and Enterprise(Depends)
“By default, content training is turned off on Organization and Enterprise plans...”

Manage AI settings and content training On the page Nov 9, 2026

Depends
Grammarly
Off for sales-bought plans only(Depends)
“For multi-user team accounts purchased directly through our website, Product Improvement and Training will initially be on, and an admin can opt out for their entire organization.”

Product Improvement and Training Control On the page Sep 25, 2026

Depends
JFrog
Anonymized usage data only(Depends)
“JFrog does not use customer artifacts or binaries to train AI language models...”

JFrog AI Transparency On the page Sep 25, 2026

Depends
Keenable
De-identified usage data(Depends)
“to use Usage Data in de-identified or aggregated form to maintain, improve, train, finetune, and enhance Keenable's current and future products, services and technologies...”

Terms of Service On the page Oct 5, 2026

Depends
Mistral AI
Paid API: you can opt out(Depends)
“Customers retain full control over this processing and have the right to opt out at any time.”

Do you use my user data to train your AI models? On the page Oct 26, 2026

Depends
OpenRouter
Depends on the provider(Depends)
“Wherever possible, OpenRouter works with providers to ensure that prompts will not be trained on, but there are exceptions.”

Provider logging and data policies On the page Nov 9, 2026

Depends
Raindrop
Only custom models for you(Depends)
“Raindrop may use Your Data to train or fine-tune machine learning or artificial intelligence models solely for your benefit and use...”

Terms of Service On the page Sep 25, 2026

Depends
Replit
No for Enterprise only(Depends)
“For Enterprise users, Replit does not use customer content to train its models.”

Model improvement On the page Oct 19, 2026

Depends
Runpod
Only aggregated, anonymized use(Depends)
“use Your Content in an aggregated and anonymized form to update and improve the Service and Runpod's related products and services.”

Terms of Service On the page Oct 19, 2026

Depends
Slack
No for gen AI; opt out of ML(Depends)
“Slack will not use Customer Data to train generative AI models unless Customer provides affirmative opt-in consent.”

Privacy Principles: Search, Learning and AI On the page Nov 2, 2026

Depends
Vercel
Paid Pro: off unless opted in(Depends)
“If you are on a paid Pro plan, Model Training is not enabled by default and you may opt in to Model Training...”

Terms of Service On the page Nov 2, 2026

Not stated on the watched pages: Ai2, Alteryx, Astral, Character.AI, Cloudera, Convex, CoreWeave, Cursor, DataBahn, Datadog, DataStax, DeepSeek, DuckDB, ElevenLabs, Fastly, Fish Audio, Fivetran, Fly.io, Groq, Groundcover, HashiCorp, Inflection AI, Instinct, John Snow Labs, Lambda, Lyzr, Meticulous, Midjourney, MongoDB, MotherDuck, Naïve, Oracle, Orchestra, Oxylabs, Pinecone, Poe (Quora), Replicate, Rillet, RWX, Samsara, SAS, Sazabi, sci2sci, Skan AI, SoundHound AI, Streamlit, Supabase, Temporal, Veridion.