This policy is part of the Vulcan Workspace Terms of Use. It applies to everyone who uses Vulcan, including a superuser and a custom Vulcan implementation, unless the agreement for that implementation says otherwise.
This policy is the full statement of the safety rules. The Terms of Use and the privacy notice point here, so that the rules are written once.
If you break this policy, we may suspend your access or close your account. See section 10.
1The rules apply to everyone
These rules hold for every account, every conversation, and every country we serve.
2Safety on Vulcan
Two rules, and they never change:
- We monitor activity. We keep a record of how Vulcan is used, and we use it to keep Vulcan and the people who use it safe.
- We look for requests that seek harm. We have systems that detect chats seeking instruction on violence or self harm. Vulcan refuses those requests.
3Review of a flagged chat
A safety system may flag a chat. A person at HammerForgeAI may read it to decide what to do. We keep the record described in section 2.
4Protect people
You may not use Vulcan to:
- threaten, intimidate, harass or defame anyone;
- promote or help with suicide, self harm, or a disordered eating pattern;
- create sexual violence, or intimate content about a person without their consent;
- promote terrorism or violence, including violence against a group of people;
- develop, obtain or use a weapon, including a conventional weapon, a firearm, or a chemical, biological, radiological, nuclear or explosive weapon;
- carry out an unlawful activity, or trade in unlawful goods or services;
- damage or break into another person’s system or property, including malicious cyber activity;
- run a real-money gambling service;
- give advice that needs a licence, such as legal or medical advice, without a licensed professional involved;
- test our safety systems without our agreement;
- get around a safeguard or a safety measure;
- work for a national security or intelligence purpose without our written approval.
5Respect privacy
You may not use Vulcan to:
- build a facial-recognition database without the consent of the people in it;
- identify a person in real time by remote biometric means in a public place;
- use a person’s likeness, including a photorealistic image or a copy of their voice, without their consent, in a way that could confuse what is real;
- judge or sort people by their social behaviour, their personal traits or their biometric data, including social scoring, profiling, or inferring a sensitive attribute;
- infer how a person feels at work or at school, unless a medical or safety reason requires it;
- predict whether a person will commit a crime from their personal traits or from a profile;
- gather, watch, profile or distribute a person’s private or sensitive information without authorisation.
6Keep minors safe
A minor is a person under 18, in every country. You may hold an account from the age of 16. The protections in this section apply to everyone under 18, and that includes an account holder who is 16 or 17.
You may not use Vulcan to:
- create child sexual abuse material, whether or not any part of it is AI-generated;
- groom a minor;
- show a minor age-inappropriate content, including graphic self-harm, sexual or violent content;
- promote unhealthy dieting or exercise to a minor;
- shame or stigmatise a minor’s body type or appearance;
- set a dangerous challenge for a minor;
- run sexual or violent roleplay with a minor;
- give a minor access to an age-restricted good or activity.
We report child sexual abuse material to the competent authorities, as the mandatory reporting law requires. We also report the endangerment of a child to the competent authorities where the law requires it, or where we judge that a child is at risk.
7Empower people
You may not use Vulcan to:
- do academic work dishonestly;
- deceive, defraud, scam, spam, or impersonate a person or an organisation;
- run a political campaign, lobby, interfere with an election, or discourage people from voting;
- make a high-stakes decision about a person without human review. This covers:
- Medical and Legal
- Financial activities and credit, and Employment
- Housing and Education
These are the areas where a wrong output is most likely to affect a person’s health, livelihood or safety. We apply this rule in good faith, and we review the list as the service develops.
You may not use Vulcan to promote or glorify hatred, racism, or discrimination against a group or a person because of who they are.
8Our own limits, stated
Output can be wrong. A model can state something false with confidence, miss something, or misread a source.
Vulcan is not a substitute for professional advice, or for a qualified engineer’s sign-off.
Do not use output as the only basis for a decision where a mistake could cause harm, loss or a safety risk.
9Tell us about a problem
A breach of this policy, or a harmful answer. Write to [email protected]. We review every report, and we may act on it.
A security vulnerability. Do not test the service without telling us first. Write to [email protected] about one.
A problem with your account. Sign in at Vulcan, or write to [email protected].
If you need this information, or a way to contact us, in another format, see the accessibility section of the Terms of Use.
10What happens if you break this policy
- We may suspend your access, or close your account.
- We may do so without notice.
- We do not offer an appeal.
- We may report an unlawful act to the competent authorities.
- Where the law gives you a right to notice or to an explanation, we will give it.
11Contact
HammerForgeAI Corp. [email protected]
If you need this policy, or a way to contact us, in another format, see the accessibility section of the Terms of Use.