AI & automation

Generative AI for NGO Campaigns: Opportunities, Swiss Legal Framework, and Responsible Governance

Strategic and legal guide for Swiss NGOs implementing generative AI: multilingual content creation, data protection (FADP), regulatory context, and human oversight.

May 9, 2025 by ewm

Generative AI can help an NGO prepare text, translations and summaries. Start with public documents and tasks whose results a person can check. Facts, beneficiary circumstances and fundraising appeals require particular care.

Context of the Swiss Charitable Sector and Innovation Opportunities

Choose a need observed in the team: drafting an outline, summarising approved information or adapting a message. Assess checking time as well as generation time. Producing more content does not demonstrate more donations.

In this evolving landscape, generative AI serves as a practical efficiency tool for communication and fundraising teams:

Campaign and Content Drafting

Generating initial outlines for awareness articles, newsletters, press releases, and social media posts.

Multilingual Communication

Efficiently adapting messages across national languages (French, German, Italian) to reach diverse Swiss regions with cultural nuance.

Segmentation and Personalization

Analyzing engagement patterns to align donor appeals with specific interest areas.

Operational Time Savings

Automating administrative summaries and information material preparation.

To assist organizations in implementing these digital workflows, specialists provide tailored EWM generative AI solutions aligned with non-profit standards.

Swiss Legal Framework: Data Protection and Intellectual Property

AI adoption by NGOs must strictly comply with Swiss legislation. Two primary legal areas govern these activities: personal data protection and copyright law.

The Federal Act on Data Protection (FADP)

In force since September 1, 2023, the revised FADP applies directly whenever AI systems process personal data. As emphasized by the Federal Data Protection and Information Commissioner during Data Protection Day, AI technology itself is neutral under the law, but data protection rules apply immediately when identifiable personal information is processed. NGOs must fulfill two key obligations:

Enhanced Transparency

Donors and users must be clearly informed when interacting with automated systems or when their personal data is processed in AI workflows.

Prompt Data Security

Personal or confidential donor details must never be entered into unsecured public AI tools.

Intellectual Property and Platform Terms

According to guidance from the Swiss Federal Institute of Intellectual Property (IPI), Swiss copyright law protects only human creations. Content produced solely by AI algorithms does not automatically qualify for copyright protection unless a human contributor makes a substantial creative contribution. Organizations must also examine platform terms of use regarding commercial rights and data usage. Detailed practical insights are available on the Swiss government's SME Portal.

Check applicable rules

Do not automatically apply EU AI Act obligations to every Swiss NGO. Assess activities, roles and relevant markets. For your 2027 plan, assign someone to monitor official sources; an announced reform is not a rule already in force.

1. Needs Assessment and Usage Scoping

The initial phase involves mapping repetitive tasks suited for AI assistance, such as draft translation or report formatting. Organizations should establish clear internal guidelines specifying approved tools and data restrictions.

2. Compliance Verification and Data Safeguards

Before using a platform, check data, contracts, retention, transfers and training settings. Minimise supplied information and avoid identifiable beneficiary records in unapproved tools. High-risk processing needs an appropriate assessment.

3. Human Review Protocol Implementation

In this method, each output intended for publication receives review by qualified staff. Human oversight corrects factual inaccuracies, identifies potential algorithmic biases, and checks that messaging remains aligned with organizational values.

4. Staff Training and Responsible AI Use

Training communication and fundraising staff on AI limitations—such as factual hallucinations or stereotyping—is essential. Programs should focus on effective prompt construction and critical evaluation of generated outputs.

5. Periodic Audits and Regulatory Monitoring

In preparation for upcoming legislative developments expected at the end of 2026 and into 2027, non-profits should regularly review internal protocols against federal guidance.

Decision Aid: Practical Application Scenario

To illustrate these principles, consider the fictional example of the foundation "Eco-Solidarity", preparing a multilingual emergency appeal following a regional weather event:

Step 1

A communications officer uses AI to generate three text outlines based on anonymized public reports of local emergency needs.

Step 2

The editorial lead selects the preferred version, verifies factual details, and adjusts the tone to ensure authentic alignment with the foundation's voice.

Step 3

The AI tool provides initial draft translations in German and Italian, which are subsequently reviewed and refined by regional translators for cultural accuracy.

Before publication

the team checks facts, rights and translations, then decides whether to send the content. The example reports neither time saved nor fundraising results and does not guarantee compliance.

To structure your AI integration strategy, request an EWM strategic consultation.