Prodigy - OpenText Legal Hold Integration and Automation

Common Integration Use Cases Between Prodigy and OpenText Legal Hold

Below are practical integration scenarios where Prodigy and OpenText Legal Hold can work together to support legal, compliance, and AI operations while reducing risk and manual effort.

1. Preserve AI Training Data Under Legal Hold

Data flow: OpenText Legal Hold ? Prodigy

When datasets used for model training become subject to litigation or investigation, OpenText Legal Hold can flag the relevant source records, files, or custodian-owned content so Prodigy users are prevented from deleting or overwriting those items. This helps AI teams continue annotation work on preserved copies while legal teams maintain defensible retention.

• Automatically identify datasets tied to held custodians or matters
• Block deletion of raw images, text, or annotation exports in Prodigy projects
• Reduce risk of spoliation during active legal matters

2. Suspend Annotation on Sensitive Content During Litigation

Data flow: OpenText Legal Hold ? Prodigy

If a dataset contains customer communications, employee records, or regulated content under legal hold, the integration can mark those items as restricted in Prodigy so annotators cannot modify labels, comments, or metadata. This ensures the integrity of evidence while preserving the rest of the workflow for non-held data.

• Apply read-only status to held records
• Prevent label changes that could alter evidentiary value
• Maintain auditability for legal review

3. Create Legal Hold from AI Data Discovery Results

Data flow: Prodigy ? OpenText Legal Hold

During model development, Prodigy may surface datasets that include sensitive or potentially relevant content, such as customer complaints, HR text, or incident images. If reviewers identify material that may be relevant to an investigation, Prodigy can send dataset identifiers, file references, or review outcomes to OpenText Legal Hold to initiate or update a hold matter.

• Convert annotation review findings into legal hold candidates
• Accelerate preservation of newly discovered relevant content
• Reduce dependence on manual handoffs between AI and legal teams

4. Track Custodian-Owned Data Used in Model Training

Data flow: Bi-directional

OpenText Legal Hold manages custodians and held content, while Prodigy manages the datasets used for training. A bi-directional integration can map custodians to the datasets they contributed, enabling legal teams to see which training assets are associated with a hold and allowing AI teams to exclude or isolate those records from active labeling queues.

• Link custodians to Prodigy projects and source files
• Identify which annotation tasks involve held content
• Support defensible reporting for legal and compliance teams

5. Preserve Annotation History and Model-Training Evidence

Data flow: Prodigy ? OpenText Legal Hold

For matters involving AI decision-making, organizations may need to preserve the exact state of labeled data, reviewer actions, and annotation timestamps. Prodigy can export annotation history, task assignments, and dataset versions to OpenText Legal Hold so legal teams can retain evidence of how training data was created and modified.

• Archive annotation logs and dataset versions
• Support eDiscovery requests involving AI development records
• Provide defensible evidence of data handling and review steps

6. Restrict New Labeling on Content Added to an Active Hold

Data flow: OpenText Legal Hold ? Prodigy

When OpenText Legal Hold adds new content to an active matter, the integration can automatically notify Prodigy and place those items into a restricted state. This prevents new labeling tasks from being assigned to content that must remain unchanged, reducing the chance of accidental alteration by data science or operations teams.

• Automatically quarantine newly held records
• Stop task assignment for affected datasets
• Keep annotation workflows aligned with legal preservation rules

7. Support Compliance Reviews for AI Data Governance

Data flow: Prodigy ? OpenText Legal Hold

Organizations often need to demonstrate that AI training data was handled appropriately during investigations or audits. Prodigy can send dataset lineage, annotation completion status, and reviewer identity information to OpenText Legal Hold, giving legal and compliance teams a centralized record of how potentially relevant AI data was processed.

• Document who labeled or reviewed specific content
• Capture dataset lineage for audit and legal review
• Improve response time for regulatory inquiries and litigation support

These integrations help legal teams preserve relevant information while allowing AI teams to continue working with clear controls, better traceability, and lower operational risk.