PathFactory AI FAQ

PathFactory AI FAQ

(Last Updated: Oct 27, 2025)

Introduction

The following is a compilation of some frequently asked questions about the use of AI in the provision of PathFactory services. 

FAQ

1. Does PathFactory use AI in its products and services?

Yes. PathFactory uses AI in several product features, including recommendations, conversations, tagging and summaries as explained below:

1. 1. • • Content Library: Content that is uploaded or linked by the customer to its Content Library is used to train the AI model. AI is used to tag the content by taxonomy, and tags can then be used to search, identify and filter content for use by buyers and marketers. 
        • Website Tools, Content Playlists, Templated Experiences and Revenue Intelligence: AI is optionally used to recommend personalized content, enabling the buyer to access and view the next best content based on their buyer journey, past content they have viewed and firmographic segment (if available). These recommendations are surfaced as follows: 
          • To buyers on the customer’s website using Website Tools;
          • To buyers via emails, paid ads, social media, etc. using Content Playlists and Templated Experiences; and
          • To sellers within customer’s CRM or PathFactory Platform using Revenue Intelligence.
        • Templated Experience Authoring: Generative AI is used to assist with translations of copy in templated experiences pages, as well as labels in templated experiences and content playlists. AI is also used to automatically generate themes, templates and landing pages based on a sample website, Figma design or style guide. Translation assistance and automated generation using AI are optional features in templated experience authoring.
        • ChatFactory: Generative AI is used to answer buyer and marketer questions based on customer’s content. It also uses AI to provide content recommendations and to automatically generate personalized microsites for buyers based on the conversation and recommended content.
        • Headless AI Services: PathFactory provides AI services that can be programmatically used by third-party activation components to provide recommendations, as well as to surface ChatFactory in preferred forms and locations.

2. What information is used by the different AI features?

• Extraction, tagging, topic modeling, keyphrase generation use content and content metadata.
• Recommendations use content, content metadata, buyer engagement information and buyer firmographic data.
• Conversations use content, content metadata and buyer questions.
• Personalized conversation summaries use conversations.

3. Do customers get their own AI instance, or are AI models shared across customers?

Each customer instance is isolated from other instances. AI models are not shared across customers. 

4. Is customer content used to train or improve PathFactory’s AI?

Customer content is not used to train PathFactory’s proprietary AI system or third-party AI. Customer content is only used to train the AI model for that customer’s specific instance in order to provide better recommendations to buyers, to better manage customer content and to provide responses to buyers when using ChatFactory. 

Contractual safeguards are in place with  PathFactory’s third-party LLM providers, to restrict them from using customer content to train their LLMs and they are not provided with any visitor identification or engagement information. See Questions 8 and 9 below for more information on third-party AI. 

5. Do customers have to expose sensitive content to the LLM in using ChatFactory?

No. Firstly, your marketing content is not typically sensitive in nature as it’s meant for public consumption. Secondly, customers have full control over the information that is exposed to the LLM. Customers create “Collections” which are curated sets of content assets that they wish to use with ChatFactory, which it then uses, to generate conversation responses. 

6. Is bias a concern in using PathFactory’s Services which involve the use of AI?

We believe that the potential for bias is very limited given the type of AI used, the context in which it is used and the controls that are in place.  Some of the AI used in our Services is “traditional AI”, which relies upon pre-programmed rules and algorithms to perform simple but specific tasks for e.g. tagging of your content by taxonomy.  Although ChatFactory and some other features involve the use of “generative AI”, they are used in the B2B marketing context with “low risk” AI decision-making that is limited to conversation responses (“outputs”) or to extract text from content, translation and code generation. These features do not have available to them or use biographical information (for e.g. sex, gender, religion, date of birth) to generate outputs for a particular visitor. Further, as discussed in the question above, the customer has complete control over the content assets that are used with ChatFactory to source responses and sets the rules, down to the asset, for determining the content recommendations that are surfaced. Customers may view a record of what content recommendations were made in PathFactory’s analytics.

7. What personal data may be processed by PathFactory in relation to its AI services? 

• IP addresses (where a visitor has provided cookie consent)
• Contact details (where PFRI is used or where a visitor has consented to cookies and completed a form with contact information)
• Conversations with the ChatFactory agent
• Visitor/Audience engagement with customer’s content 

8. Is PathFactory’s AI proprietary, or does it use a third-party service? 

PathFactory has two AI systems in place: 

1) A Core AI system: proprietary, core to using PathFactory capabilities, isolated for each client (i.e. customer data is not shared across customers, nor used to train the AI system for other customers);  and

2) A Generative AI system: leverages third party LLM, optional for use but impacts ability to use certain products / features. 

Further details are provided below: 

9. What third-party LLMs are used?

PathFactory’s Generative AI uses the following LLMs:

• • Anthropic’s Claude 3 Sonnet
  • Open AI’s GPT-4o, GPT-4o-mini, GPT-4.1, GPT-5
  • Meta’s Llama 3.3

10. Are additional AI-powered features planned in the future?

PathFactory plans to continue adding AI-powered features in the future. These include the use of AI to:

• • provide explanations and insights to analytics
  • easily build templated experiences and content playlists
  • enrich content with thumbnails, summaries and other useful metadata
  • act as copilots for Sales and Marketing to provide personalized microsites to their customers, accounts and industry segments
  • enrich PathFactory’s activation components

11. Does PathFactory host LLMs?

No. PathFactory does not host or manage any LLMs. Instead, PathFactory makes calls to third-party LLMs whenever LLM capabilities are required. 

12. What is the security of data in transit and at rest?

PathFactory understands that customers need to partner with companies that they can trust with their data.  PathFactory is SOC-2 Type II compliant for all data.

Views: 113