Our Acceptable Use Policy (also referred to as “AUP”) aims to ensure that our products and services are being used safely in a manner that does not create new risks for our customers or society. It applies to anyone who uses Gray Swan’s products and services in any way. We may employ automated or manual means to verify that your use of the products and services is consistent with this AUP, and take steps to suspend or terminate your access to the products and services if your use is deemed to violate it.
Usage Standards
Gray Swan’s products and services should never be used to harm someone or otherwise jeopardize the safety of any person or animal, violate anyone’s privacy or autonomy, damage someone’s reputation, create or distribute false or misleading information, engage in hateful or discriminatory behavior, or in any way further an illegal, fraudulent, abusive, or predatory activity or goal.
Do not harm anyone
Harming or jeopardizing the safety of a person or animal encompasses a wide range of actions that could cause physical, emotional, or psychological harm. This includes, but is not limited to, using our products and services to:
- Plan, coordinate, or encourage acts of violence or terrorism
- Design, manufacture, or distribute weapons or explosives
- Provide instructions for creating harmful substances or devices
- Incite self-harm or suicide
- Engage in cyberbullying or harassment, online or otherwise
- Develop or spread misinformation about health and safety practices
- Interfere with critical infrastructure or public safety systems
- Exploit, manipulate, or abuse vulnerable populations
- Generate, spread, or promote content that depicts or otherwise facilitates child sexual abuse, sex with minors, nonconsensual sex, sextortion, or sex trafficking
Do not violate peoples’ privacy or autonomy
Violations of privacy or autonomy involve actions that infringe upon an individual's right to control their personal information, make independent decisions, or maintain their personal boundaries. Examples include, but are not limited to:
- Collecting, storing, or sharing personal data without explicit consent
- Engaging in unauthorized surveillance or monitoring of individuals
- Bypassing security measures to access private accounts or protected information
- Creating deceptive fake likenesses (“deepfakes”) or manipulated media
- Impersonating individuals to gain access to personal or private information
- Generating phishing attempts or social engineering attacks
- Automating mass collection of publicly available personal data
- Inferring or revealing sensitive personal information
- Manipulating decision-making processes through assisted persuasion
Do not damage anyone’s reputation
Damaging someone's reputation involves taking actions that unfairly or maliciously harm an individual's or organization's standing in their community or profession. Examples of such activities include, but are not limited to:
- Creating or spreading false or misleading information about a person or entity
- Generating fake reviews or testimonials to harm a business or professional
- Creating or amplifying defamatory content
- Manipulating any type of media to portray someone in a false or unflattering light
- Impersonating individuals or organizations to make damaging statements in their name
- Automating the spread of rumors or unverified negative information
- Using our products and services coordinate smear campaigns
- Generating misleading information to damage someone’s credibility
- Using our products and services in attempts to mimic real people for deceptive purposes
Do not create false or misleading information
Creating or distributing false or misleading information, often referred to as misinformation or disinformation, involves the deliberate spread of inaccurate or deceptive content. Examples of creating or distributing false or misleading information include, but are not limited to:
- Generating fake news articles or fabricated stories
- Creating fake or manipulated media that aims to misrepresent events
- Producing content that presents false scientific claims or medical advice
- Using our products and services to generate, amplify, or spread conspiracy theories
- Generating false financial reports or market analyses
- Fabricating product reviews or testimonials
- Creating chatbots or AI personas that spread misinformation
Do not engage in hateful or discriminatory activities
Our goods and services must not be used to promote, engage in, or facilitate hateful behavior or discrimination of any kind. This includes actions that target individuals or groups based on race, ethnicity, national origin, religion, sexual orientation, gender identity, disability, age, or any other protected characteristic. Examples of prohibited hateful or discriminatory activities include, but are not limited to:
- Generating or spreading content that: promotes or employs hate speech or extremist ideologies, denies or minimizes atrocities or violent acts, or reinforces harmful stereotypes or biases
- Creating chatbots or AI personas that express hateful or discriminatory views
- Creating or deploying mechanisms unfairly discriminate in hiring, lending, or other decision-making processes
- Targeting, profiling, or excluding specific groups in advertising or service provision based on protected characteristics
- Coordinating harassment or disinformation campaigns that target marginalized groups
Do not engage in illegal activities
Using our AI platform to engage in or facilitate illegal, fraudulent, abusive, or predatory activities is strictly prohibited. These activities encompass a wide range of behaviors that exploit, deceive, or harm others for personal gain or malicious purposes. Examples include, but are not limited to:
- Using the products and services to assist with financial fraud
- Generating fake legal documents or credentials
- Using the products and services for assistance with gaining unauthorized access to computer systems or networks
- Using the products and services to assist the creation of counterfeit goods or services
- Using the products and services to assist with evading law enforcement or regulatory oversight
- Using the products and services to assist with the production or distribution of illegal or highly-regulated materials and substances
The categories and examples given above are meant to illustrate the aim and intent of our AUP, and are not exhaustive. Use of our products and services that falls within the purview of these categories may result in immediate suspension or termination of access to the products and services. We reserve the right to review usage that falls outside of this scope, and take appropriate action.
Disclosure Requirements
Users of our products and services are required to disclose the involvement of AI in generating content or responses in certain situations. This disclosure ensures that individuals who interact with AI-generated content do so knowingly, and can make informed decisions. Disclosure is mandatory in the following circumstances:
- When our products and services are used to generate responses in customer service interactions.
- When our products and services are used to provide analysis and recommend outcomes in certain professional settings, including healthcare, legal services, housing, employment, lending, accounting, and financial advice.
- When our products and services are used in an interactive setting where the human participant may otherwise believe that they are interacting with another person, for any reason.
Disclosure is not required when the interaction or publication of content generated by our products and services is wholly mediated by a person who is able to edit responses and ensure compliance with our AUP.
Otherwise, users must provide clear and conspicuous disclosure that AI has been used to generate or contribute to the content. Failure to disclose AI involvement in these contexts may be considered a violation of our AUP and could result in suspension or termination of access to our services.
