Guidelines for the use of House of Control's AI services

When using our AI services, you are not permitted to do any of the following, either directly or indirectly:

Use, or allow other end users to use, generated data to develop models that compete with any of the services or software offered by HoC or HoC's subcontractors, or deconstruct any of the services, software, or their models (or extract any components of the foregoing)
Performing or contributing to dangerous, illegal, or malicious activities, including, but not limited to:
- Facilitating or promoting illegal activities or violations of the law, such as:
  - Promoting or generating content related to child abuse or exploitation
  - Promoting or facilitating the sale of, or providing instructions for synthesizing or accessing, illegal drugs, goods, or services
  - Aiding or encouraging users to commit any form of crime
  - Promoting or generating violent extremism or terrorist content
Abuse, disrupt, or otherwise harm services (or enable others to do so), such as:
- Promoting or facilitating the generation or distribution of spam
- Generating content for deceptive or fraudulent activities, scams, phishing, or malware.
Attempting to override or circumvent security filters or deliberately causing the model to act in a way that violates our guidelines.
Generating content that may harm or promote harm to individuals or a group, such as:
- Generating content that promotes or incites hatred
- Seeking ways to harass or bully others in order to threaten, abuse, or insult them
- Generating content that contributes to, promotes, or incites violence
- Generating content that contributes to, promotes, or incites self-harm
- Generating personally identifiable information for distribution or other harm
- Tracking or monitoring people without their consent
Generating malware, including code designed to damage computer systems, provide unauthorized access to computer systems, or render computer systems inaccessible.
Generate content that may have unfair or adverse consequences for people, particularly consequences related to sensitive or protected characteristics.
Generating and distributing content with the intent tomisinform, misrepresent, or mislead, including:
- Misrepresenting the origin of generated content by claiming that the content was created by a human, or representing generated content as original works in order to mislead
- Generating content that impersonates an individual (living or deceased) without explicit disclosure, in order to mislead
- Misleading or false claims of expertise or capabilities, particularly in sensitive areas (e.g., health, finance, public services, or legal services)
Making automated decisions in areas that affect material or individual rights or welfare (e.g., finance, legal matters, employment, health, housing, insurance, and welfare social services), including high-impact decision-making in the public sector such as investigation, law enforcement, migration, and asylum.
Generate sexually explicit content, including content created for purposes related to pornography or sexual gratification (e.g., sexual chatbots).