Which privacy dimensions does this tool compare?

It compares whether the provider trains on your inputs by default, how long data is retained, whether data is shared with third parties, what enterprise/API protections exist, and the provider's stated GDPR position. These are the dimensions that most often decide whether a provider is acceptable for business use.

Are consumer and API/enterprise terms different?

Yes, often dramatically. Many providers train on free consumer chat data by default but contractually exclude API and enterprise traffic from training. The tool notes where the consumer tier and the business tier diverge so you don't apply the wrong assumption.

Is the data in this tool authoritative?

No — it is a summarized, generalized snapshot to help you shortlist and ask the right questions. Provider terms change frequently. Always confirm the current, signed terms on the provider's own legal pages before relying on them for a compliance decision.

Does opting out of training stop all data processing?

No. Opting out of model training typically still leaves operational processing, abuse-monitoring retention, and logging in place. Training opt-out and zero data retention are separate settings, and not every provider offers both.

Does this tool send my selections anywhere?

No. The comparison runs entirely in your browser. Nothing about which providers you compare is uploaded.

What is the AI Provider Terms Comparison Tool?

Select two or more AI providers and compare their terms on key privacy dimensions — training data use, retention, sharing, enterprise protections, and GDPR compliance — in a clear side-by-side table you can read at a glance. It runs free in your browser on Gera Tools, with nothing uploaded.

AI Provider Terms Comparison Tool

Name: AI Provider Terms Comparison Tool
Creator: Gera Tools
License: https://creativecommons.org/licenses/by/4.0/

Choosing an AI provider is partly a privacy decision: does it train on your prompts, how long does it keep them, and will it stand behind GDPR? This tool puts the leading providers side by side on the dimensions that actually decide whether a model is safe to use with real business or customer data.

Why privacy terms differ so much between providers

AI providers operate across radically different product tiers, and the privacy terms often differ as much between tiers as between providers. The free consumer tier of one product and the enterprise API of a competitor can offer more similar privacy postures than the free and paid tiers of the same company. Understanding those differences is what lets you make a defensible procurement decision rather than an assumption.

The five dimensions this tool compares are chosen because they are the ones that actually decide whether you can put real business data into a system:

Training on your data — whether your inputs are used to improve the model. This is the single most-asked question in enterprise procurement and the one most often misunderstood.
Retention period — how long inputs and outputs are stored, and whether that differs between free and paid tiers.
Third-party sharing — whether data flows to sub-processors for moderation, annotation, or service operation, and what controls exist.
Enterprise/API protections — whether API and enterprise plans carry a data-processing agreement, contractual exclusions from training, and zero-retention options.
GDPR posture — whether the provider offers a signed DPA, names a legal basis for processing, and has a EU or UK representative.

How it works

Pick two or more providers. The tool builds a comparison table across those five privacy dimensions with a short, plain-language summary in each cell. Where a provider’s consumer tier differs from its API or enterprise tier, the cell calls that out, because the difference is frequently the whole story.

The two distinctions that catch organisations out

Training opt-out ≠ zero retention. Switching off model training on your data means your inputs are not used to improve future model versions. It does not mean they are deleted immediately. A provider may retain inputs for days or weeks for abuse monitoring, safety review, or debugging, even with training opted out. If you need short or zero retention, that is a separate and usually premium setting — confirm it explicitly.

A DPA is the only GDPR artifact that counts. A provider’s website saying “GDPR compliant” is a marketing statement, not a legal commitment. A signed Data Processing Agreement (Article 28 GDPR) is what makes the relationship lawful for EU personal data — it names the sub-processors, sets instructions for processing, and assigns liability. Without it, relying on a provider for EU personal data is a compliance gap, regardless of their public claims.

Tips and notes

Training opt-out ≠ zero retention. They are separate switches. Confirm both if you need both.
A DPA is the deciding artifact for EU data — a marketing claim of “GDPR compliant” is not the same as a signed Data Processing Agreement.
This table is a shortlisting aid, not legal advice. Always verify the live terms on the provider’s own legal pages, and pair this with the Data Residency Checker for region requirements.
Re-check at renewal. Privacy terms change. A policy reviewed at procurement may look different at renewal, especially for rapidly evolving AI products.