A public catalog of AI for the public good
Where could AI actually help?
The AI Opportunity Inventory identifies, catalogues, and analyzes AI use cases with real potential to advance major societal goals — from healthcare and education to energy and disaster response — and subjects each one to legal and policy scrutiny it can be trusted on.
16
Opportunities logged
10
Societal domains
9+
Partner institutions
Opportunity, documented — not press-released.
Policymakers, researchers, and funders struggle to find credible AI use cases outside of company press releases and blog posts. Every AI application carries the possibility of positive and negative social outcomes; this inventory deliberately focuses on the opportunity for positive results — and welcomes scrutiny of every entry, starting with our own.
By systematically mapping AI uses against public policy problems, the inventory aims to foster collaboration, direct resources effectively, and inform policy that accelerates innovation for the public good.
How it works
01
Submit
Researchers, civil society organizations, government bodies, and AI labs submit detailed AI use cases: the problem addressed, the system used, who benefits, the evidence, and the barriers to adoption.
Express interest ↗02
The Raw Opportunity Bank
Every submission lands in a publicly viewable bank — an unfiltered look at the incoming landscape, exactly as submitted, open for independent analysis and collaboration.
Browse the bank →03
Researched & published
Student researchers put promising entries under real legal and policy scrutiny — short research memos, built with structured coaching and reviewed by human editors, then published as posts anyone can read.
See how memos are built →
For student researchers
Have an idea? Here's the path.
1
Verify you're a member
On the assistant page, enter the email you joined with. Not a member yet? The interest form takes two minutes.
2
Pitch your idea
Describe the AI opportunity in a sentence or two. The assistant immediately searches everything in the bank — and every idea another student has already claimed — so you know within a minute whether yours is open.
3
Develop the memo, with a guide
If your idea holds up, it's recorded under your name and you get a structured path: a memo template, staged feedback on each section, and a coach that demands real sources — and won't write it for you.
4
Editors take it from there
When your draft meets the bar, the assistant submits it to the human editors, who decide what gets published.
Full program details at law.utexas.edu/ai.
The Research Assistant
A coach for the research,
not a ghostwriter.
Anyone can ask it what's in the collection and what it all means. Students and researchers in the program get more: it tests whether your idea is genuinely new, pushes you to back every claim with real sources, and hands your memo to human editors when it's ready. It asks the hard questions — it won't write the memo for you.
Open the assistantFrom the workflow
Decision: Variant — revise direction
Confidence: 0.72
Closest match: Library-augmented generation for scholarly research
Required next step: Narrow to court-record collections and identify the operating institution.
Suggested memo focus: Access-to-justice deployment context; records privacy as the lead legal issue.
Research an opportunity.
Shape the record.
Students and researchers at participating institutions can join the community to submit AI opportunities, write research memos, or serve as editors.
Join the community ↗