AI Content & Third-Party Provider Disclosure
PRISM coordinates third-party AI systems. Outputs require human review and may not be unique, accurate, or legally protectable.
Last updated: September 22, 2026
On this page
1. AI-generated content
PRISM outputs are generated or transformed by artificial-intelligence systems. Results are probabilistic and may contain errors, artifacts, bias, offensive material, unexpected similarities, or factual inaccuracies. Preview and review every output before relying on, publishing, or distributing it.
2. Provider processing
Prompts, settings, and relevant uploaded media may be transmitted to the provider selected by you or by PRISM Auto to perform the requested generation. Current or supported providers may include OpenAI, Google, Runway, Luma, Higgsfield, ElevenLabs, HeyGen, Recraft, Ideogram, and future providers.
Provider processing, retention, safety, and training practices vary according to provider, endpoint, account configuration, and applicable service terms. PRISM does not promise that every provider uses identical policies. PRISM itself does not claim to train a proprietary PRISM model on customer content unless a future feature specifically says so and provides any required notice or consent.
3. PRISM Auto
PRISM Auto may choose a provider and model using requested medium, quality, capabilities, availability, performance, and estimated cost. Automatic selection does not guarantee the lowest cost, fastest result, or best subjective result. Where manual selection is available, PRISM respects that selection unless fallback is authorized or required to complete the requested workflow.
4. PRISM Verify
PRISM Verify performs automated technical checks intended to identify conditions such as missing files, unreadable formats, implausible dimensions or duration, empty output, or other detectable generation failures. A passed result is not legal clearance, copyright clearance, trademark clearance, factual verification, authenticity certification, safety certification, quality guarantee, or confirmation that content is suitable for publication or commercial use.
5. Rights and originality
Rights in AI-generated material vary by jurisdiction and circumstances. PRISM does not guarantee that output is copyrightable, exclusive, original, patentable, trademark-clear, publicity-right-clear, non-infringing, or commercially usable. Similar or identical outputs may be generated for others. Obtain legal advice when rights clearance matters.
6. People and voices
You are responsible for securing authorization required to use a person’s face, body, likeness, image, voice, biometric information, performance, or identity. Do not use generated or altered media for impersonation, fraud, non-consensual sexual content, harassment, fabricated evidence, false endorsement, or materially deceptive civic or commercial communication.
7. High-impact use
Do not use output as the sole or determinative basis for high-impact decisions involving employment, housing, credit, insurance, education admission, healthcare access, legal services, essential government services, or law enforcement. AI output is not professional advice.
8. Availability
Providers, models, capabilities, duration limits, quality levels, pricing, and availability may change or be discontinued. PRISM does not guarantee permanent access to any named provider or model.
9. Your responsibility
You remain responsible for your prompts, uploads, settings, selected models, outputs, disclosures, and downstream use. Follow the Terms of Service and Acceptable Use Policy and disclose synthetic or altered media where required by law, platform rules, contract, or context.