This Code of Conduct sets behavioural standards for everyone who participates in SAIN activities, and defines what it means to be part of Safe AI Netherlands (SAIN).
1. Our Values
- Respect. We treat every person with dignity, regardless of background, expertise, or viewpoint. We engage with ideas on their merits.
- Intellectual honesty. We represent evidence fairly, acknowledge uncertainty, distinguish personal views from established consensus, and correct ourselves when wrong.
- Inclusivity. We actively welcome people of all backgrounds, disciplines, and experience levels. AI safety benefits from diverse perspectives, technical and non-technical, academic and professional, junior and senior, different cultures, genders, etc.
- Collaboration. We default to cooperation. We share knowledge, credit contributions, and support each other's growth.
- Safety. We are committed to both AI safety and interpersonal safety. Every SAIN space must be one where people feel secure enough to learn, question, and contribute.
2. Scope
This Code applies to all SAIN-organised events (national and local, in-person and online), all SAIN digital spaces (WhatsApp, Discord, email lists, social media), all activities conducted under the SAIN name (courses, research collaborations, hackathons, publications, public appearances), and all individuals acting in a SAIN capacity (team members, volunteers, course participants, event attendees, research collaborators, and speakers). It applies equally across all chapters.
3. Expected Behaviour
- Engage constructively. Disagree with ideas, not with people. Offer feedback that aims to improve, not diminish.
- Act in good faith. Assume positive intent. Give people the benefit of the doubt, especially when perspectives differ.
- Welcome newcomers. Everyone starts somewhere. Make space for people beginning their AI safety journey.
- Respect diverse perspectives on AI safety. SAIN's main focus is on risks from transformative AI, especially existential risks. However, we want everyone to engage respectfully with the full spectrum, from near-term algorithmic harms to existential risks, from technical alignment to governance and policy. Engage respectfully across this spectrum, even when you personally prioritise one area.
- Give proper credit. Acknowledge contributions in research, content, and organisational work.
- Protect confidentiality. Do not share personal information, private communications, or unpublished research without consent.
- Respect privacy and consent. Only photograph, record (audio or video) or share screenshots of other people or of private SAIN spaces and conversations with their consent. Do not share someone's private information, such as their home address or phone number, without their consent.
- Use SAIN's name responsibly. Do not make public statements on behalf of SAIN without authorisation. When expressing personal views, make clear they are your own.
4. Unacceptable Behaviour
The following are not tolerated in any SAIN space or activity:
- Harassment: stalking, intimidation, deliberate misgendering, sustained disruption of discussions.
- Sexual harassment (seksuele intimidatie): any form of verbal, non-verbal or physical behaviour with a sexual connotation that has the purpose or effect of violating a person's dignity, in particular when a threatening, hostile, offensive, degrading or hurtful situation is created. Examples include sexual remarks, jokes or innuendo; intimate questions; staring or sexual gestures; showing sexual or pornographic images (including via Discord or other messaging); unwanted touching; and continuing to contact someone after they have said no.
- Discrimination: treating someone unfavorably based on race, ethnicity, nationality, gender, gender identity, sexual orientation, disability, age, religion, or any other protected characteristic.
- Bullying: personal attacks, public shaming, threats, and patterns of behaviour designed to make someone feel unwelcome or unsafe.
- Misrepresentation: claiming to speak for SAIN without authorisation, or using the SAIN brand to endorse personal or external agendas.
- Retaliation: any adverse action against someone for reporting a concern or participating in a conduct process.
- Privacy violations: photographing, recording (audio or video) or sharing screenshots of other people or of private SAIN spaces and conversations without their consent, and sharing someone's private information (such as their home address or phone number) without their consent.
5. Research and Professional Integrity
SAIN is a place where people build research track records and careers, so honest work matters. The following are breaches of SAIN's integrity standards and are handled through the process in Section 6:
- Academic dishonesty: plagiarism, data fabrication, or deliberate misrepresentation of evidence.
- Undisclosed conflicts of interest: failing to disclose a conflict of interest that is relevant to SAIN decisions, research, recruitment or partnerships.
6. Reporting and Enforcement
Anyone who experiences or witnesses a violation is encouraged to report it:
- Local chapter leadership: for issues within a specific chapter. Contact your chapter's Chapter Director or Community Manager.
- SAIN national: for issues involving chapter leadership, cross-chapter matters, or situations where the chapter team is not sufficient or local reporting feels inappropriate. Contact: info@safeainetherlands.org.
Reports can be made in writing or in person. Anonymous reports are accepted, though they may limit investigative capacity.
Process: (1) Acknowledgment within 2-3 working days. (2) Assessment of severity; local matters are handled by the chapter with SAIN national guidance if needed, serious or cross-chapter matters by SAIN national. While a report is being assessed or investigated, the person reported may be asked to step back from SAIN activities until the process is complete; this is a precaution, not a finding of wrongdoing. (3) Investigation where warranted. All parties heard, confidentiality maintained. (4) Resolution, which may range from a private conversation to a formal warning, temporary suspension, or permanent removal from SAIN activities. The Chapter Directors discuss and decide on the outcome, unless the National Team needs to be involved. Either party may ask for the outcome to be reviewed within 14 days; the review is carried out by someone who was not involved in the original decision (for chapter-level matters, a member of the National Team).
Principles: Consequences are proportional to severity and any pattern of past conduct. All parties are treated with respect. The accused has the opportunity to respond. Details are shared only with those who need to know. SAIN actively protects anyone who reports in good faith. Personal data in reports is handled in line with SAIN's Privacy Policy.
7. Commitment to Intellectual Diversity
AI safety is a broad and evolving field. Reasonable people disagree about the relative severity of different risks, the best mitigation approaches, and timelines. SAIN does not impose a single orthodoxy.
SAIN's primary focus is on skilling up talent and creating the pipeline that allows people to work on AI risks. We are concerned with risks arising from increasingly capable and potentially transformative AI systems, which include risks that could have profound consequences for humanity's long-term future. Members need not agree on every claim, forecast, or prioritisation, but should engage seriously and in good faith with the challenges posed by transformative AI.
What we require:
- Respect for the mission. SAIN focuses on the risks of transformative AI. We are also concerned by the potential existential risks that could emerge from increasingly capable systems. Members are expected to respect this focus, even if they differ on specific priorities, timelines, or interventions.
- Evidence-based discourse. Arguments grounded in evidence and reasoning, not appeals to identity, authority, or tribal affiliation.
- Constructive disagreement. Serious disagreements about strategy, governance, technical approaches, or priorities are normal and valuable. We encourage respectful debate aimed at improving understanding.
- Openness to updating. Intellectual humility is expected. The field moves fast, evidence changes, and people should be willing to revise their views.
- No gatekeeping. Anyone engaging in good faith with the mission belongs here.
Organisational Neutrality and Individual Expression
SAIN's role is to build the infrastructure, opportunities, and connections that help people contribute to AI safety. We bring people into the field, develop their capabilities, and connect them with research, projects, organisations, and other opportunities to do meaningful work. In practice, this means:
- We enable the field; we do not represent every position within it. SAIN may share opportunities, events, and resources from organisations across the AI safety landscape (e.g., linking to a PauseAI event, an Anthropic talk, or an AI governance workshop) without this constituting an endorsement of that organisation's specific positions.
- We connect people to opportunities across the field. We may share events, research opportunities, fellowships, jobs, and other resources from organisations with different approaches to AI safety. Our role is to help people find opportunities to learn and contribute.
- Individuals within SAIN are free to hold and express specific opinions on AI safety matters, including publicly. What they may not do is present those opinions as official SAIN positions (see Section 3: "Use SAIN's name responsibly").
This distinction must be especially clear to chapter Communications teams and anyone communicating on SAIN's behalf. When in doubt, frame it as: "SAIN member X believes..." rather than "SAIN believes..."