Responsible AI Campaign
Ethical, accountable AI in social services
GASWT's signature campaign — closing the gap between how fast AI tools are being adopted in social services and how slowly professional standards for using them are catching up.
AI-assisted tools are already used across child welfare, benefits eligibility, risk assessment and case triage in social services worldwide — often adopted faster than the professional standards for using them responsibly. The Responsible AI Campaign is GASWT's push to close that gap: research, published standards, and direct pressure on the institutions and vendors deploying these tools.
Four standards we're pushing for
Bias & harm review
AI tools used in eligibility, risk-scoring or case-assignment decisions should be independently reviewed for bias before deployment — and re-reviewed as they're retrained.
A human in the loop
No AI system should be the final word on a decision that affects someone's access to protection, custody or care. A qualified person makes the call, informed by the tool.
Explainability
People affected by an AI-assisted decision have a right to a plain-language explanation of how it was reached — not just a score.
Data minimization
Systems should collect only the data a decision genuinely requires, held only as long as it's needed — not everything a vendor's model could theoretically use.
Get involved
Research volunteers, technologists and social workers all have a role in this campaign — from reviewing tools already in use to co-authoring the standards themselves.