Data use and AI improvement

Your information, responsibly used

Our products improve through use, helping us deliver better results for your agency and across our product family.

We may use lawfully available public business information, business content we are authorised to use, and genuinely anonymised information to develop, train, test and improve our AI, products and services. This applies across our present and future products as our technology and methods evolve.

Business customers grant us a non-exclusive licence, to the extent they hold the necessary rights, to use their non-personal business content for those purposes through authorised suppliers. Customers retain their existing rights. Our reusable models, methods and improvements remain ours or our licensors’, and may continue to be used after the service ends, subject to confidentiality, third-party rights and applicable law.

We do not use identifiable personal data about you, your team or your clients to train or improve our own wider AI models. Public availability alone does not make information non-personal: staff names, photographs, social profiles and information identifying property occupants can still be personal data.

Information used for anonymous learning must first be processed so that individuals cannot reasonably be identified, including by combining it with other available information. Removing names or encrypting data alone is not anonymisation. Any processing needed to anonymise personal data must itself be lawful and, where we act as a processor, covered by documented customer instructions. This agreement does not authorise unrestricted access to private records.

Your products can use authorised agency, team and property information to provide accurate answers and deliver your agreed services. This is separate from wider model training. Creating a requested personal avatar or synthetic voice remains subject to that individual’s separate, informed authorisation for the specified service; it does not give us permission for wider model training or other uses of their identity.

We handle personal data in accordance with applicable data-protection law and protect it with appropriate security measures, including encryption and access controls. Confidential information remains protected, and private agency records are not made available to other customers.

We use commercial technology suppliers under their applicable standard terms and data-use policies. Depending on the service, these permit storage, security review and improvement of their services or models, including use of de-identified or aggregated information. We explain material supplier processing and available choices for the selected feature, including a real person’s likeness or voice before creation. We remain responsible for choosing and using suppliers lawfully, protecting confidential information and respecting your rights. Our commitments about training our own models do not imply identical data-use or retention practices at every supplier.

Public sources and supplied content remain subject to copyright, database rights, platform conditions and other applicable rights. An agency’s permission extends only to rights it can grant. We do not treat a public website or social profile as unrestricted permission to reuse personal information or protected content.

The permitted purposes above are not limited to a particular model, supplier or technical method. We may evolve those methods within these purposes and protections without asking for a new permission for each technical change. Required supplier notices, transfer safeguards and contractual change procedures still apply.

This policy applies to ModelProp services identified in your agreement, including services coordinated through OmniProp and the Console. It does not purchase or activate another product, extend access to another agency or group company, replace a separate presenter permission, or retrospectively change an existing contract. Existing agreements continue until validly varied.