Impact
Neutral has never meant equal.
Not on maps, not in tech, and not in markets.
Women in high-income countries are nearly 3x more likely to hold jobs vulnerable to AI.
Female engineers using AI were rated 9% less competent than male peers for identical work.
First Prompt conducts and elevates research on women's relationships with AI. We also track how women and other participants feel about AI before and after our programs, both to keep the work high quality and to learn from their feedback. That evidence feeds everything we do, from workshops to advisory to policy advocacy, helping create a more equitable map of the AI transition.
What the data shows
Delivered in Boston, Belfast, rural Uganda and beyond, with independently validated, statistically significant gains in confidence, familiarity, and perceived usefulness of AI.
High confidence using AI tools
AwakenHub · Belfast · 27 participants
Self-rated confidence
WILD Virtual AI Masterclass · 73 assessed before, 30 after
Preparedness to manage AI risk
WILD Virtual AI Masterclass · 73 assessed before, 30 after
Mountaintop International
A dependent-samples t-test run by Dr. Linda Lin (Professor of Psychology, Emmanuel College) found statistically significant, large-effect-size improvements in familiarity, confidence, and perceived usefulness of AI after the workshop. The largest gain was in confidence, replicated across two independent cohorts. Every participant who answered the recommendation question said they would recommend it.
Imagine Her
93% of staff were already using generative AI before training, mostly on free tools with no data protections. First Prompt ran a full discovery-to-measurement engagement: staff survey, a written AI usage policy aligned with Uganda's data protection law, hands-on training for the entire organization, and a post-session confidence assessment.
"I left the session feeling more confident about using AI to improve my work. She made me realize the decision to use AI should come from a point of power while also knowing its limits."
"This is a life line I was looking for and needed."
"I just assumed that everything AI produced was gospel. I didn't recognize that there are biases and hallucinations."
Methodology: pre- and post-session self-assessment, collected via form and live polling immediately before and after each session. Sample sizes vary by program and by question; figures above are participant averages, not projections.
Stay up to date
The latest thinking on women, gender and AI, plus periodic updates from First Prompt.
Subscribing happens on Substack. Their privacy policy applies to anything you enter there.