# Good Practices

This section defines the mandatory policies, guidelines, and standards that govern all data and AI projects at UNESCO. Developed collaboratively by Digital Business Solutions (DBS), the Data Protection Office (DPO), the Cybersecurity Unit, and Legal Affairs, these practices ensure every initiative aligns with UNESCO's values, international standards, and obligations around privacy, security, ethics, and open access.

***

#### 🪪 Data Quality

All datasets produced or consumed by UNESCO projects must meet defined quality standards covering completeness, consistency, accuracy, and documentation.

* [UNESCO Data Quality Standards and Guidelines](/unesco-data-ai/good-practices/unesco-data-quality-standards-and-guidelines.md)

***

#### 🔏 Privacy & Personal Data Protection

Projects handling personal data must comply with UNESCO's data protection principles and consult the Data Protection Office at the design stage.

* [UNESCO's Principles on Personal Data Protection and Privacy](https://www.unesco.org/en/privacy-policy)
* Contact the Data Protection Office (DPO) *(internal)*

***

#### 🛡️ Cybersecurity

All infrastructure must follow UNESCO's cybersecurity policies, including approved cloud hosting configurations and access management practices.

* [Hosting in Azure — Microsoft Cloud Adoption Framework](https://learn.microsoft.com/en-us/azure/cloud-adoption-framework/ready/azure-best-practices/resource-abbreviations)
* SSO & Data Access — Data Governance *(internal wiki)*
* Contact the Cybersecurity Unit *(internal)*

***

#### 🕊️ AI Ethics & Guardrails

AI systems must be developed and deployed in accordance with UNESCO's Recommendation on the Ethics of AI, with guardrail mechanisms to prevent harmful outputs.

* [UNESCO Recommendation on the Ethics of Artificial Intelligence](https://www.unesco.org/en/articles/recommendation-ethics-artificial-intelligence)
* AI Chat Guardrail Mechanism
* Ethics Considerations — Harmful Opinions 
* Environmental Impact of AI 

***

#### ⚖️ Data Licensing

UNESCO data and code outputs are published under open access principles. Projects must apply appropriate licensing and consult Legal Affairs when in doubt.

* [UNESCO Access to Information Policy](https://www.unesco.org/en/unesco-access-information-policy)
* [Creative Commons Attribution-ShareAlike 4.0 (CC BY-SA 4.0)](https://creativecommons.org/licenses/by-sa/4.0/)
* Contact Legal Affairs 

***

#### 🗺️ Cartography & Borders

All maps published by UNESCO must use the official UNESCO tile background to ensure compliance with UN cartographic standards and border representations.

* [UNESCO Geospatial Disclaimer](https://www.unesco.org/en/geospatial/disclaimer?hub=85601)
* Default tile URL for public dataviz tools: `http://www.unesco.org/tiles/clearmap/{z}/{x}/{y}.png`

***

#### 🌐 Open Source

UNESCO adopts open source as the default approach for data and AI projects, in line with UN system principles and UNESCO's own Open Science mandate — reducing vendor lock-in, ensuring auditability, and reflecting the diversity of global AI development.

**UN System Foundation**

The [UN Open Source Principles](https://unite.un.org/en/news/osi-first-endorse-united-nations-open-source-principles), adopted by the [UN Chief Executive Board's Digital Technology Network](https://unsceb.org/) and endorsed by the [Open Source Initiative (OSI)](https://opensource.org/), establish that UN projects should be *open by default*, *designed for reusability*, and built to *sustain and scale* across the UN system.

**UNESCO Open Science Alignment**

UNESCO's [Open Science](https://www.unesco.org/en/open-science) framework promotes equitable access to knowledge, open data, and open infrastructure as global public goods — open source software practices are a direct extension of this commitment.

**Strategic Imperatives**

* **Diversity of AI models**: Preference for openly accessible models — including those from non-Western institutions and the global research community (e.g. [Hugging Face](https://huggingface.co/)) — prevents dependence on a small number of dominant providers.
* **Vendor independence**: Open standards and open source components preserve UNESCO's ability to self-host, switch providers, and avoid long-term lock-in.
* **Auditability & ethics**: Open systems enable independent review of AI outputs, supporting compliance with the [UNESCO AI Ethics Recommendation](https://www.unesco.org/en/articles/recommendation-ethics-artificial-intelligence).

**Guidelines for Project Teams**

* Publish code under [OSI-approved licenses](https://opensource.org/licenses) or [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)
* Contribute improvements back to upstream open source communities
* Prefer open, auditable AI models and frameworks unless security or functional requirements justify otherwise
* Document projects in line with the [UN Open Source Principles](https://unite.un.org/en/news/osi-first-endorse-united-nations-open-source-principles)


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://unesco.gitbook.io/unesco-data-ai/good-practices.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.