AI Privacy Statement - Freedom Scientific

AI Privacy Statement


Freedom Scientific acts as a pass-through to Open AI’s Chat GPT and Anthropic Claude Enterprise Services. We, as well as our partners, do not store or train on any images that are sent through Picture Smart or screenshots of controls sent through the UI Labeler. We pass this content to either Chat GPT or Claude using the associated prompt only. In order to enhance the quality of our documentation, we retain queries and feedback provided by users through FSCompanion. For information about how Chat GPT and Claude process your information, please see the Open AI Enterprise Privacy page and the Commercial Terms of Service \ Anthropic.

View our full Complete Data Privacy Policy.

Contact Us

If you have questions about the privacy aspects of our Services or would like to make a complaint, please call 1-800-444-4443, ext. 1191 or email us at [email protected].

Privacy and Security FAQ

Page Explorer and FSCompanion

Q: What is the source of the AI model utilized with FSCompanion and Page Explorer?
A:
We use OpenAI’s enterprise LLM as the foundation, and we customize its behavior using Vispero’s proprietary data and system-level prompts.

Q: Do you collect any of the website content submitted to Page Explorer for analysis?
A: No. We do not collect or store any of the website content that the user submits for analysis.

Q: Do you save the generated answer from Page Explorer?
A: To improve Page Explorer, we temporarily store the generated response as part of the FSCompanion conversation data, along with any follow-up questions asked through the interface.

Q: What data is collected?
A:
FSCompanion temporarily stores the text of user queries and user-submitted feedback to help improve the product experience. If available, we also collect general location information (such as country or region) to better understand where our tools are being used. This data is approximate and not linked to individual users.

Q: Do you store any audio recordings?
A:
Audio from the FSCompanion dictation feature is not saved, only the transcribed text is stored for telemetry.

Q: How long is user collected data stored in your database?
A:
Collected telemetry data is stored for a maximum of two years.

Q: Can you confirm that customer-submitted queries are not used for model training purposes?
A:
Yes. The enterprise API from OpenAI utilized by Freedom Scientific is designed to prevent the storage or use of customer-submitted queries for model training. Customer questions are only sent to OpenAI for processing, such as transcribing audio dictation.

Q: Can you verify the security of the communication channel between FSCompanion and the OpenAI model?
A:
Yes, all communication between FSCompanion and OpenAI is conducted over TLS, ensuring end-to-end secure communication.

Picture Smart AI and AI Based UI Labeler

Q: What is the source of the AI models utilized with Picture Smart AI and the AI Based UI Labeler?
A:
Freedom Scientific leverages publicly accessible models from Anthropic and OpenAI, specifically through their enterprise-level APIs. While requests from Picture Smart AI are sent to both models, requests from the AI Based UI Labeler are sent only to OpenAI.

Q: Can you confirm that customer-submitted image data is not used for model training purposes?
A:
Yes. The enterprise APIs from Anthropic and OpenAI utilized by Freedom Scientific are designed to prevent the storage or use of customer-submitted images for model training.

Q: Specifically describe how an image is processed within your system.
A:
An image is processed by transmitting a combination of pre-defined system prompts, user-generated prompts, and configurable parameters like temperature to the respective API.

Q: How is image data transferred from the JAWS application to your service, and from what infrastructure are the API calls executed, including what security measures are in place?
A:
Image data is transmitted directly from the JAWS application to the Freedom Scientific service. API calls originate from a secure Freedom Scientific backend, authenticated via access key management.

Q: Detail the security protocols employed for communication between the JAWS application and your web service.
A:
All communication with the Freedom Scientific web service is secured using TLS connections over HTTPS (port 443).

Q: What data retention policies are in place for image data within your service?
A:
The Freedom Scientific service maintains a temporary cache of image data and associated responses to support image-based chat functionality.

Q: What is the duration of this temporary cache, and how is data purged?
A:
The cache is automatically cleared every four hours. No images are permanently stored.

Q: Can you verify the security of the communication channel between your service and the Anthropic/OpenAI models?
A:
Yes, all communication between the Freedom Scientific service and Anthropic/OpenAI models is conducted over TLS, ensuring end-to-end secure communication.

Q: Is it possible to disable the Picture Smart AI functionality during the setup?
A:
Yes, Picture Smart AI can be disabled during installation. See: Software Installation Guide