University students develop artificial intelligence applications for mental health, economics and bias detection

In recent months, student projects at several U.S. universities have shown how artificial intelligence can tackle practical challenges in mental health, bias detection, and economics education. A University of San Francisco program had students from 13 majors build AI-powered wellness tools, while San Francisco State University recognized original student work in its first Student AI Awards. At a national competition in Austin, a team from Morgan State University won a $35,000 grand prize for an app that helps college students manage social anxiety.

Wellness and sustainability take shape in campus projects

The University of San Francisco's measured embrace of generative AI led to a program where students developed solutions centered on sustainability, health, and technology. One environmental science major created an application to identify plants and trees. Other participants built a well-being tracker and a website generator that helps parents and lower-income households find safe, educational online content. The university's president had emphasized using AI to serve human needs, and these projects gave that principle a concrete form.

Across higher education, the integration of AI tools is accelerating, with programs like the University of San Francisco's wellness solutions initiative showing the potential of AI for Education. The student work demonstrates that ethical adoption doesn't mean avoiding the technology-it means directing it toward specific, human-centered outcomes.

Student awards highlight bias detection and event management

San Francisco State University's inaugural Student AI Awards in May honored work that combined innovation with critical analysis. One software engineering project examined how AI systems can perpetuate systemic biases, finding that models trained on historical data may reproduce patterns of prejudice and exclusion. Another award-winning entry streamlined hackathon management with a platform that improved submission, judging, and event operations. A third project delivered a human-centered AI prototype for an industry challenge, stressing that technology includes people, processes, and implementation-not just automation.

The university participates in the SUGAR network, which connects students with industry challenges and reinforces that broader perspective. This approach treats AI development as a discipline that requires scrutiny of its social effects alongside technical skill.

SocialSense AI wins top prize at HBCU competition

Four student innovators from Morgan State University's Human-AI eXperience (HAX) Lab earned first place and $35,000 at the 2026 PROPEL Center Future of Tech Innovation Challenge. The team beat out 1,100 applicants from 89 HBCUs, with only 70 finalists advancing to the national stage. Their winning project, SocialSense AI, is a privacy-focused application built within the Apple ecosystem that helps college students manage social anxiety and build confidence in interpersonal interactions.

The platform uses data from iPhone and Apple Watch to analyze conversational dynamics and deliver personalized feedback, actionable insights, and progress tracking. Sensitive data stays stored on the user's device, a privacy-centric design that judges praised. Team members described the experience as an opportunity for interdisciplinary collaboration that showed the tangible impact of emerging technologies.

Economics students turn monetary policy into music

At North Carolina Agricultural and Technical State University, associate professor Scott Simkins is bringing generative AI into economics education. Since 2022, he has co-led a faculty learning community on responsible AI use. After attending the university's 2025 Artificial Intelligence Conference, he designed an assignment where students used ChatGPT for lyric generation and Suno for music production to create original songs explaining the Federal Reserve's monetary policy challenges.

The projects integrated macroeconomic models and covered topics such as the economic impact of government shutdowns and Federal Open Market Committee debates. The resulting songs translated complex concepts into accessible formats, and Simkins' pedagogy was later featured on Marketplace, a nationally syndicated business and economics program. He continues to encourage colleagues to weave generative AI into their teaching.

Why this matters for education professionals

These examples offer a clear pattern: when faculty give students structured opportunities to apply AI to real problems, the results go beyond novelty. The projects described here produced working tools, surfaced bias in training data, and made abstract economic policy memorable. For educators, the takeaway is that AI assignments work best when they are tied to specific, measurable outcomes-whether that's a functioning app, a competition entry, or a song that explains the federal funds rate. The technology is not a replacement for critical thinking but a medium that demands it.