Alignment with humanitarian principles

  • 4 minutes

When nonprofits use AI, the goal is to support people, not to put them at risk. To do this, AI must follow three key humanitarian principles that protect rights, strengthen trust, and prevent harm.

Dignity: Protect people and their data

Dignity means respecting every person and keeping their information safe. Nonprofits often collect sensitive data, so any AI system must handle it carefully.

This requires:

  • Being transparent about how data is collected and used.
  • Getting informed consent whenever possible.
  • Using strong protection tools (anonymization, encryption).
  • Making sure AI doesn't cause harm or put anyone at risk.
  • Safeguarding the privacy of stakeholders and communities served when using their data in AI tools.

This is part of the basic "Do No Harm" principle.

Accountability: Stay responsible for AI decisions

Nonprofits, not machines, are responsible for AI's impact.

Accountability means:

  • Keeping AI systems transparent (not "black boxes").
  • Setting clear policies and processes around AI use.
  • Reviewing AI outputs for accuracy, fairness, and safety every time.
  • Allowing people to challenge or question AI-generated decisions.
  • Ensuring people remain in control of important decisions.

AI should support staff, not replace their judgment.

Fairness: Avoid bias and treat people equally

AI shouldn't reinforce inequalities or disadvantage certain groups.

Key questions to check fairness:

  • Is the data used to train the AI system representative of all groups we serve?
  • Could the AI unfairly affect a specific community?
  • Can we audit or review how the system makes decisions?

Nonprofits must actively look for unfair patterns and take steps to correct them because AI can learn hidden biases from its training data.

Put these principles into practice

Nonprofits should build these principles into their existing governance and data processes to use AI responsibly. This includes:

  • Clear ethical guidelines
  • Legal and data protection requirements
  • Regular audits and oversight
  • Strong cybersecurity
  • Transparent communication about how AI is used

Tools like NetHope's Humanitarian AI Code of Conduct help organizations ensure their AI systems are safe, fair, and aligned with humanitarian values.