OpenAI
· OpenAI Usage Policies
This is a hard, unconditional restriction that applies to every user and operator without exception; violation would constitute both a policy breach and potentially criminal conduct under laws in most jurisdictions.
These restrictions define the outer boundary of permitted Bedrock use cases and carry account termination risk for violations; organizations providing Bedrock-based services to end users must ensure their downstream use cases comply with these restrictions.
This provision places affirmative compliance obligations on the deploying organization rather than on DeepSeek, meaning that legal responsibility for lawful and non-harmful use of the model's outputs rests with the user or deploying entity.
These prohibitions apply to all licensees and flow downstream to end users of products built on self-hosted models, meaning deployers are responsible for enforcing these restrictions within their own platforms.
Canva
· Canva Privacy Policy
This provision establishes the operational scope of Canva's AI/ML development activities and clarifies that user-generated content constitutes input data for model improvement. The clause conditions such use on existing contractual rights under the Terms of Use and applicable legal authority, creating a framework for distinguishing permitted from restricted training applications.
This clause defines the conditions under which Google's content delivery obligation terminates and establishes the exclusive remedies available when removal occurs. The refund-as-sole-remedy limitation restricts the scope of potential claims to the purchase price, excluding other damages or relief.
This provision directly constrains how developers and businesses may deploy Gemini in automated workflows, and the requirement for 'appropriate human oversight' creates an operational standard that may require compliance documentation.
Shein
· Shein Terms and Conditions
The provision establishes the operational infrastructure through which Shein shares user activity data with external advertising partners for behavioral targeting and campaign optimization. This multi-network approach enables coordinated data collection across multiple advertising ecosystems.
The clause authorizes Together AI to incorporate user-submitted content into service operations and product development without additional compensation or licensing restrictions. The sublicensable and transferable language permits Together AI to authorize third parties to exercise these same rights.
The clause establishes LinkedIn's operational rights to deploy user content across its service ecosystem, including through sublicensing arrangements with third parties. The transferable and sublicensable language permits LinkedIn to delegate these usage rights to other entities without additional user authorization.
The provision establishes Jasper's operational rights to incorporate user content into service improvements and to distribute that content through third-party partnerships without requiring per-use compensation or separate authorization from the user for each distribution instance.
The provision establishes Cohere's operational authority to incorporate user-generated inputs into service operations, product development, and distribution without requiring separate compensation or individual consent for each use case. The sublicense right enables Cohere to authorize third parties to exercise the same rights over user content.
Runway
· Runway Terms of Service
This clause establishes the scope of rights Runway acquires over publicly posted user content, determining what uses of that content the platform is permitted to make without separate compensation or consent beyond the initial submission.
This clause establishes Wix's operational authority to exploit user content across all forms and media without time limitation or geographic restriction. The non-exclusive structure permits users to retain rights while simultaneously authorizing Wix to exercise the full bundle of copyright-related permissions without compensation.
Notion
· Notion Terms of Service
The BAA creates a formal compliance framework that obligates Notion to implement administrative, physical, and technical safeguards for PHI, establish breach notification procedures, permit audit and inspection rights for the covered entity, and ensure that any subcontractors also comply with HIPAA requirements. This provision is operationally significant because it establishes the legal basis under which healthcare organizations can lawfully use Notion's platform to store or process patient health information.
This provision establishes the outer boundary of permissible use and, because it is framed as a strict prohibition, violation would constitute a per se breach of the Terms of Service regardless of other approvals obtained.
Pika
· Pika Acceptable Use Policy
This provision prohibits child exploitation content regardless of whether it is prohibited by law in a given jurisdiction, and commits Pika to mandatory reporting to NCMEC, which is consistent with obligations under 18 U.S.C. Section 2258A for electronic service providers.
This clause establishes categorical restrictions on service use related to child safety by prohibiting a defined set of activities and content types. The provision operates as a use restriction that Anthropic enforces through its API governance framework.
The provision defines X's operational framework for identifying and removing child sexual exploitation material from its platform, establishing institutional obligations and enforcement mechanisms for compliance with applicable legal requirements.
The provision defines the scope of data collection by the AI assistant service and identifies what constitutes collectible Prompt Data. This establishes the operational basis for the service to process and retain development environment content submitted during AI interaction.
The clause establishes the scope of OpenAI's operational rights to process and utilize user content as part of service delivery and product development. This licensing structure permits OpenAI to incorporate user interactions into model training and service improvement without additional compensation or per-use licensing fees.
The provision defines the operational framework through which the service maintains child safety compliance, establishes the authority to implement automated and manual content controls, and sets parameters for user conduct on the platform. This establishes binding obligations for users to comply with moderation requirements and allows the service to enforce safety policies through content restrictions.
The clause creates distinct intellectual property regimes tied to payment status, determining whether users hold copyright to outputs or operate under a non-commercial license from Midjourney. This directly affects permitted use cases, commercialization rights, and asset disposition across the user base.
The clause establishes the scope of regulatory protections available to different account types. SIPC insurance provides coverage up to $500,000 per customer in the event of broker insolvency, but this protection does not apply to cryptocurrency positions, which means account holders bear the risk of loss in the absence of broker protection for those assets.
This provision reflects a legal obligation under US federal law (PROTECT Act, FOSTA-SESTA adjacent obligations) and similar laws globally, and violations carry criminal liability independent of platform enforcement.
This is the most absolute provision in the policy, combining mandatory law enforcement referral with automatic account termination, distinguishing it from other violations where enforcement is discretionary.
OpenAI
· OpenAI Usage Policies
This provision applies to all users and operators and covers generation of offensive cyber tools, though the document implicitly acknowledges a distinction between prohibited offensive tool creation and permitted defensive security research — a distinction that may not always be clear in practice.
This prohibition applies to all users and API operators, and covers both direct creation of malicious software and providing technical assistance that enables cyberattacks causing significant harm.
This provision addresses one of the most socially significant risks of AI voice technology, specifically the generation of synthetic audio designed to manipulate public opinion or deceive audiences, and it has direct implications under emerging synthetic media election laws.
Suno
· Suno Terms of Service
The clause establishes boundaries on permissible use of Service outputs and restricts reverse engineering, competitive development, and technical circumvention of access controls. This defines the scope of authorized use and protects the Service's content protection mechanisms and geographic licensing structure.