VYSUAL LABS LLC

AI Processing & Asset Recognition Policy

Last Updated: March 14, 2026

This AI Processing & Asset Recognition Policy ("AI Policy") explains how Vysual Labs LLC ("Vysual," "we," "us," or "our") uses artificial intelligence and automated processing technologies within the Vysual platform (the "Platform").

This AI Policy is incorporated into and subject to the Vysual Terms of Service. In the event of a conflict between this AI Policy and the Terms of Service, the Terms of Service shall control. Capitalized terms not defined here have the meanings given in the Terms of Service.

This policy is designed to answer three questions plainly: what AI does to your content, who does it, and what they are prohibited from doing with it.

1. What AI Processing Means for Your Content

When you upload assets to the Platform, Vysual uses automated and AI-assisted processing to generate features that help you organize, search, and use your media library. This processing happens at the infrastructure level — you do not need to do anything for it to occur, and it begins as soon as an asset is uploaded.

AI processing on the Platform may include:

Transcription

Audio and video files are analyzed to generate text transcripts. Transcripts are indexed to make your content searchable by spoken word. Transcripts may contain errors, particularly for technical terminology, accents, background noise, or low-quality audio.

Visual Recognition and Tagging

Images and video frames are analyzed to identify objects, scenes, activities, and other visual attributes. These are attached as searchable tags to your assets. Visual recognition results are probabilistic and may be incomplete or incorrect.

Metadata Extraction and Indexing

File-level metadata (format, duration, dimensions, codec, creation date) is extracted from uploaded files and indexed for filtering and search. This is deterministic processing, not AI inference.

Thumbnail and Preview Generation

The Platform automatically generates thumbnail images and preview clips from video and image files to support the asset browsing experience.

Asset Categorization

Based on file type, visual attributes, and extracted metadata, assets may be automatically categorized to assist with library organization. Categorization suggestions can be reviewed and overridden by users at any time.

2. What Vysual Will Not Do With Your Content

The following restrictions apply to Vysual and to all third-party AI vendors acting on Vysual's behalf:

Hard prohibition: Vysual will not use the content of your uploaded assets — including video, audio, images, or documents — to train, fine-tune, benchmark, or otherwise develop any AI or machine learning model, whether operated by Vysual or a third party, without your separate, express written consent.

This prohibition is absolute and applies regardless of whether the assets are processed in identified or de-identified form. The content of your media library is yours. It is processed to serve you, not to build products for others.

Additionally, Vysual will not:

• sell, license, or otherwise transfer access to your uploaded assets or AI-generated metadata to third parties for their own commercial purposes
• use AI processing results from your assets to generate outputs delivered to other customers without your consent
• use biometric data derived from your assets (such as facial geometry or voiceprints) for any purpose other than generating searchable tags within your own Workspace

Vysual may use de-identified and aggregated platform-level data — such as usage patterns, feature adoption rates, error rates, and file type distributions — to improve Platform functionality. This data does not include the content of your assets or any information that could identify your organization or your users.

3. Accuracy and Limitations of AI Outputs

AI-generated outputs including transcripts, tags, labels, and categorizations are probabilistic by nature. They are provided to assist with asset organization and discovery — they are not authoritative records and should not be relied upon for compliance, legal, medical, or other high-stakes purposes without independent human review.

Known limitations include:

• Transcription accuracy varies based on audio quality, speaker accents, technical vocabulary, and background noise. Accuracy rates are not guaranteed.
• Visual recognition may misidentify objects, people, or scenes, particularly in low-resolution, partially obscured, or ambiguous images.
• AI tagging reflects the capabilities and training of the underlying third-party model at the time of processing. Tags are not updated retroactively when models are improved.
• AI outputs may reflect biases present in the training data of the underlying models. Vysual does not control the training data of third-party AI vendors.

Users are responsible for reviewing, correcting, and managing AI-generated metadata within their Workspaces. Vysual does not guarantee the accuracy, completeness, or fitness for any particular purpose of AI-generated outputs.

4. Third-Party AI Vendors

Vysual uses third-party vendors to provide certain AI processing features. Each vendor is contractually prohibited from using your asset content to train their own models and is required to implement appropriate security measures consistent with Vysual's obligations under the Terms of Service and Privacy Policy.

Current AI processing vendors are listed below. Vysual will update this table when material changes to AI vendors occur and will provide thirty (30) days' advance notice via email for any change involving vendors that process uploaded asset content.

5. Biometric Data

Certain AI processing features — including facial detection in images and video, and voice analysis in audio files — may generate or process biometric data as defined under applicable law, including the Illinois Biometric Information Privacy Act (BIPA) and similar state statutes.

Vysual's position on biometric data:

• Biometric identifiers or information derived from uploaded assets are used solely to generate searchable tags and metadata within the uploading user's own Workspace.
• Vysual does not store biometric identifiers as stand-alone data points separate from the asset record.
• Vysual does not sell, lease, trade, or profit from biometric data.
• Vysual does not use biometric data derived from one customer's assets in connection with any other customer's Workspace or for any purpose unrelated to providing the Platform to the uploading customer.

If your use of the Platform involves uploading content that contains biometric information about individuals — including video footage of employees, clients, or members of the public — you are responsible for ensuring that your collection and processing of that biometric data complies with applicable law, including obtaining any required consents from the individuals depicted.

Illinois note: If your organization is subject to BIPA or if you upload content containing faces or voices of Illinois residents, consult with legal counsel before enabling visual recognition or transcription features. BIPA provides for statutory damages of $1,000 to $5,000 per violation, per person.

6. Opt-Out from AI Processing

AI processing features are enabled by default for all Workspaces. Workspace administrators may opt their Workspace out of AI processing features by submitting a written request to support@vysual.com with the subject line "AI Processing Opt-Out" and identifying the Workspace.

Upon receiving a valid opt-out request, Vysual will:

• disable AI processing for the specified Workspace within five (5) business days
• confirm the opt-out in writing to the requesting administrator
• not retroactively process previously uploaded assets with AI features after opt-out

Opting out of AI processing will disable the following features for the affected Workspace: automated transcription, visual tagging, AI-powered search, and asset categorization suggestions. Core Platform features including storage, manual organization, and sharing will continue to function normally.

To re-enable AI processing after an opt-out, submit a written request to support@vysual.com with the subject line "AI Processing Opt-In."

Granular Opt-Out: Visual Recognition and Voice Processing

Customers who wish to disable visual recognition (facial detection, object recognition, scene analysis) or voice processing (transcription, speaker identification) specifically — without disabling all AI features — may request a granular opt-out by contacting support@vysual.com with the subject line "AI Feature Opt-Out" and specifying which feature(s) to disable and for which Workspace.

This granular opt-out is available to all customers and is particularly relevant for organizations whose uploaded content may include biometric data — such as video footage of employees, clients, or members of the public — where applicable law requires prior consent before automated biometric processing occurs. Disabling visual recognition or voice processing will not affect other AI features such as metadata extraction, thumbnail generation, or text-based search indexing.

Vysual will confirm the granular opt-out in writing within five (5) business days of receiving a valid request.

7. Acceptable Use of AI Features

Users may not use the Platform's AI features to:

• generate, process, or distribute content that violates applicable laws or regulations
• infringe the intellectual property rights of any third party
• process content containing harmful, illegal, or exploitative material
• impersonate individuals or organizations in a deceptive manner
• circumvent consent requirements for the collection or processing of biometric data
• generate outputs intended to deceive, defraud, or manipulate others

Vysual reserves the right to suspend access to AI features for any Workspace that violates these restrictions, consistent with the suspension and termination provisions of the Terms of Service.

8. Coming Soon Features

The following AI-assisted features are in development and not yet available on the Platform. This policy will be updated prior to their launch to address any additional processing activities they involve:

• AI-powered tagging and search (enhanced semantic search)
• Upload Request with automated intake processing
• Auto transcription with speaker identification

Users will be notified via email and presented with updated policy terms before any new AI feature that processes asset content in a materially different way is enabled for their Workspace.

9. Policy Updates

Vysual may update this AI Policy as AI technologies, vendor relationships, and applicable law evolve. For material changes — particularly changes that expand the categories of AI processing applied to user assets or add new AI vendors — Vysual will provide at least thirty (30) days' advance notice via email before the changes take effect.

Continued use of AI-powered features after the effective date of an updated AI Policy constitutes acceptance of the revised policy. Users who do not agree to material changes may opt out of AI processing using the process described in Section 6, or may terminate their account under the Terms of Service.

10. Contact

Questions regarding this AI Policy, requests to opt out of AI processing, or concerns about how AI features have processed your content may be directed to:

For opt-out requests, use subject line: "AI Processing Opt-Out"
For general AI policy questions, use subject line: "AI Policy Inquiry"

© 2026 Vysual Labs LLC. All rights reserved.