Migliaccio & Rathod LLP is investigating whether HubSpot customers’ CRM records, communications, files, and other business data may have been used to train HubSpot’s artificial-intelligence models through a setting that was enabled by default.
HubSpot states that it may use Customer Data to train and improve its own AI models. That data may include information from CRM systems such as contacts, deals, communications, and other customer content. HubSpot also states that some models may be trained using data across multiple customer accounts to improve AI-powered features throughout the platform.
HubSpot’s documentation states that the AI Model Training setting is turned on by default. Only users with Super Admin permissions can turn it off.
What We Are Investigating
Businesses may use HubSpot to store commercially sensitive information concerning customers, prospects, sales, contracts, communications, and internal business activities.
Migliaccio & Rathod LLP is investigating whether HubSpot customers understood that this information could also be used to improve AI models that support features across the HubSpot platform.
We are investigating whether:
- customers were clearly informed that AI Model Training was enabled by default;
- existing customers were affirmatively asked before their data became eligible for AI training;
- the disclosures adequately explained what categories of Customer Data could be used;
- customers were given a meaningful opportunity to opt out before training occurred;
- HubSpot used customer data supplied under earlier contracts or privacy disclosures for new AI-training purposes; and
- customers could meaningfully withdraw previously used data after opting out.
HubSpot states that turning off AI Model Training stops an account’s eligible customer data from being used for future training. However, HubSpot also states that it is not possible to delete previously used data from trained models, meaning the opt-out operates prospectively.
HubSpot separately states that third-party AI service providers are not permitted to use HubSpot Customer Data to train their own models. This investigation concerns HubSpot’s own use of customer data for model training.
Potential Claims May Include:
- unfair or deceptive trade practices;
- breach of contract;
- breach of the implied covenant of good faith and fair dealing;
- unjust enrichment;
- failure to adequately disclose a new use of customer data; and
- claims involving the use of commercially valuable customer information for AI development.
Signs You May Be Affected:
- Your business uses or previously used HubSpot;
- your company stores CRM records, customer communications, files, or other business information in HubSpot;
- you discovered that AI Model Training was turned on in your account;
- no one at your company remembers affirmatively enabling the setting;
- you were unaware that Customer Data could be used to train HubSpot models;
- you later turned the AI Model Training setting off;
- you learned that opting out would not remove information already used in trained models;
- your business used HubSpot before the current AI-training disclosures were introduced; or
- you have older HubSpot contracts, account notices, emails, or screenshots describing how your data could be used.
If your company used HubSpot and you learned that customer data may have been used to train HubSpot AI models through a default-on setting, we would like to hear from you.
Please complete the contact form on this page, send us an email at [email protected], or give us a call.