OpenAI applies data filtering to reduce personal information from training data and implements a fingerprinting-based opt-out system for DALL-E 3 images that extends to the GPT-4o training dataset, removing opted-out images from the training corpus.
This analysis describes what OpenAI's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
This provision establishes a disclosed data minimization practice and an operational opt-out mechanism for image training data that the document states was applied to the GPT-4o series, creating a documented basis for evaluating OpenAI's data governance practices relative to applicable privacy frameworks.
Interpretive note: The document describes personal information reduction as a filtering process without specifying scope, completeness, or how it applies across all data modalities including audio and video, creating uncertainty about the full extent of the practice.
The document states that OpenAI uses filtering processes to reduce personal information from training data and that images opted out via the DALL-E 3 mechanism were fingerprinted and removed from GPT-4o series training datasets, providing a disclosed mechanism for image opt-out that applied to this model generation.
Cross-platform context
See how other platforms handle Training Data Filtering and Personal Information Reduction and similar clauses.
Compare across platforms →"We use advanced data filtering processes to reduce personal information from training data. Upon releasing DALL-E 3, we piloted a new approach to give users the power to opt images out of training. To respect those opt-outs, we fingerprinted the images and used the fingerprints to remove all instances of the images from the training dataset for the GPT-4o series of models.Excerpt from OpenAI's GPT-4o System Card (PDF)
(1) REGULATORY LANDSCAPE: This provision engages GDPR's data minimization principle under Article 5, the right to object to processing under Article 21, and the right to erasure under Article 17, as well as CCPA's right …
Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.
Search "[your state] attorney general consumer complaint" to find your state's direct complaint form
Get the research letter
Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.
This provision establishes a disclosed data minimization practice and an operational opt-out mechanism for image training data that the document states was applied to the GPT-4o series, creating a documented basis for evaluating OpenAI's data governance practices relative to applicable privacy frameworks.
The document states that OpenAI uses filtering processes to reduce personal information from training data and that images opted out via the DALL-E 3 mechanism were fingerprinted and removed from GPT-4o series training datasets, providing a disclosed mechanism for image opt-out that applied to this model generation.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by OpenAI.