Behind HubSpot AI

The cards in this website promote transparency around HubSpot’s use of AI. Each card provides details about HubSpot’s AI-powered features. For example, learn about inputs, outputs, how your data moves through our systems and when the data may reach our trusted third-party AI service providers. Note that our AI service providers access your data for content moderation purposes and to comply with legal obligations. We do not allow our AI service providers to use your data to improve or train their models.

By being transparent, we aim to help customers better understand how our AI systems interact with their data.

Below is an explanation of the card layout that you can expect to see for each feature.

AI Product Specifications

Model Description

A description of the AI feature and the value it’s intended to bring to customers.

How to Enable/Disable

Instructions on how to enable or disable the AI product for your account.

Data used to provide the services

A list of the data used from your account or outside sources for this feature.

Does the AI system process PII?

An explanation of whether personal data from your account is used for the AI product.

Data Retention & Deletion

An explanation of how input, output, and usage data is stored and deleted.

Is your data used to train the AI system?

An explanation of whether your data is used to train the AI system that powers the feature.

Model Provider and Type

An explanation of whether HubSpot or a third-party AI service provider provides the model.

Can the user see or review all of the user inputs and outputs?

An explanation of whether the user is able to see the output of the model before any actions are taken.

Model Input

A description of the inputs we send to the model, including any account data that is sent.

Model Output

A description of the outputs we receive from the model.

List of the hub tiers that have access to the AI product.

Limitations

While AI products can be valuable tools to boost productivity, it’s important to understand that AI outputs are not perfect, for a number of reasons. The performance of AI systems is based on the data they have been trained on, which can lead to errors or inaccuracies when presented with new or unexpected situations. HubSpot’s developing AI products may provide incorrect, biased, or misleading responses. Avoid sharing any sensitive information in your prompts and check AI-generated outputs for accuracy.

HubSpot is committed to reducing the negative impact of AI limitations by focusing on trust, transparency, and accountability. Our AI Ethics principles provide a framework that informs AI products development. We strive to be transparent with our customers and partners about how we develop and deploy AI across our organization.

Note that information provided in this website may change as we update our AI products.