Community Guidelines

A space built on respect, inclusion, and shared purpose.

Every Language Matters is a global community of contributors, volunteers, researchers, and partners united by a belief that every language deserves a place in AI. These guidelines exist to keep our community safe, productive, and welcoming for everyone — regardless of background, language, or experience level.

Respect Every Voice. Honour Every Language.

Our core values

Everything we do and every interaction in our community is guided by these foundational principles.

Respect & dignity

Every person in our community deserves to be treated with kindness and respect. We celebrate difference and hold each other to a high standard of human decency.

Inclusion & belonging

We actively welcome people from all linguistic, cultural, and professional backgrounds. No language or community is too small to matter here.

Integrity & honesty

We do our work truthfully and transparently. Accurate annotation, honest feedback, and genuine collaboration are the foundation of everything we build together.

Collaboration & community

We are stronger together. We share knowledge generously, support fellow contributors, and approach every interaction as an opportunity to build something greater than ourselves.

Quality & accountability

We take our work seriously. Every annotation, review, and contribution carries responsibility — to the languages we represent and the communities who will benefit from this data.

Safety & trust

Our community is a safe space. We protect contributor privacy, handle data responsibly, and maintain an environment where everyone feels secure to participate and contribute.

Community rules

These rules apply to all members across every platform, channel, and project space operated by Every Language Matters.

1 Treat all community members with respect
Engage with fellow contributors, volunteers, and staff with courtesy and patience. Disagreements are natural — personal attacks, condescension, and dismissiveness are not. Critique ideas, never people. This applies in all written communication including task feedback, forum posts, direct messages, and reviews.
2 Contribute accurate and honest work
All annotations, transcriptions, and translations must reflect your genuine linguistic knowledge and best effort. Do not rush through tasks, fabricate outputs, or submit work outside your language competency. If you are unsure about a task, flag it rather than guess. The integrity of the datasets we build depends entirely on the honesty of each contributor.
3 Respect cultural and linguistic diversity
Our community spans over 85 countries and hundreds of languages. Do not demean, mock, or trivialise any language, dialect, accent, or cultural practice. Avoid imposing your own linguistic norms on others' contributions. What is correct in one dialect or region may differ in another — approach all language differences with curiosity and humility.
4 Protect privacy and handle data responsibly
Never share, distribute, or discuss confidential project data, partner information, or personally identifiable information outside of authorised channels. Treat all data you encounter during tasks as sensitive. Do not screenshot, copy, or store project materials beyond what is required to complete your assigned work.
5 Communicate constructively and clearly
When providing feedback on others' work, be specific, fair, and constructive. Explain your reasoning. Avoid vague or dismissive comments. When raising concerns or disagreements, use calm and measured language. Remember that many of our contributors communicate in a language that is not their first — extend patience and good faith at all times.
6 Follow all task and project instructions
Read and follow all guidelines provided for each task carefully before beginning. If instructions are unclear, ask before proceeding. Do not deviate from annotation schemas or introduce your own conventions without approval. Consistent adherence to task instructions is what makes our datasets reliable and useful for AI development.
7 Represent yourself honestly
Only apply for roles and tasks that match your genuine language skills and qualifications. Do not misrepresent your fluency, experience, or credentials. Creating multiple accounts, impersonating others, or providing false information during onboarding is a serious violation and will result in immediate removal from the platform.

What is not allowed

The following behaviours are strictly prohibited and will result in disciplinary action up to and including permanent removal from the community.

Harassment & bullying

Targeting, intimidating, threatening, or persistently harassing any member of the community in any form — public or private, direct or indirect.

Hate speech & discrimination

Using language or content that demeans, excludes, or attacks individuals or groups based on race, ethnicity, language, religion, gender, sexuality, disability, or any other characteristic.

Deliberate misinformation

Intentionally submitting incorrect annotations, fabricating translations, or introducing false linguistic data into any dataset or project workflow.

Unauthorised data sharing

Sharing, leaking, selling, or distributing any project data, partner information, or internal materials outside of approved channels or without explicit written permission.

Spam & self-promotion

Flooding community channels with unsolicited promotional content, repeated irrelevant messages, or using community platforms to advertise unrelated products or services.

Impersonation & fraud

Creating multiple accounts, impersonating another contributor or staff member, or providing false credentials, qualifications, or identity information at any stage.

Reporting

We take all reports seriously. Here is how to report a violation and what happens next.

How to report a violation

If you witness or experience behaviour that violates these guidelines, please report it promptly to our community team. Include as much detail as possible — screenshots, dates, and context all help us act quickly and fairly.

community@everylanguagematters.com

What happens after you report

Our community team reviews every report within 48 hours. Reports are handled confidentially. We investigate thoroughly and communicate outcomes to the reporting party where appropriate. All decisions are made fairly and without bias.

Enforcement

These are the steps we may take when a violation is confirmed.

Enforcement steps

Verbal warning

A private notice explaining the violation and expected behaviour going forward.

Formal warning

A written warning logged to the member's account with a clear outline of consequences.

Temporary suspension

Access to tasks and community spaces is paused pending a review period.

Permanent removal

Permanent ban from all Every Language Matters platforms, projects, and channels.

Severe violations — including hate speech, data leaks, and fraud — may result in immediate permanent removal without prior warning steps.

Together we build something worth protecting.