Juan Zhai Receives NSF CAREER Award to Improve Reliability of AI-Assisted Software

Content

Manning College of Information and Computer Sciences (CICS) Assistant Professor Juan Zhai has received a $689,531 National Science Foundation (NSF) CAREER Award to develop new ways of ensuring that software created with AI behaves as intended.  

AI tools can accelerate productivity and make software creation more accessible, but they can also introduce errors. According to Zhai's proposal, one recent study found that developers using AI coding assistants introduced 41% more bugs than developers working without them. 

“AI is changing how software is built,” she said, “but the key question is still the same: Does the software do what we meant it to do? I see formal specifications as a bridge between human intent and AI-generated code. They can help make AI-assisted development more precise, reliable, and trustworthy.”  

Zhai’s research aims to make AI-generated software more reliable by helping coding systems better capture how software should behave and ensure that the code they produce satisfies those requirements. 

Her proposal describes formal specifications as the: “the semantic anchor that allows tools and agents to reason meaningfully about behavior.” She will develop a framework that connects plain-language descriptions, source code, and precise behavioral rules, building a new dataset that links code, documentation, and formal specifications across entire software repositories. As software evolves, the system will automatically evaluate and update those specifications, helping developers verify that AI-generated code remains aligned with intended requirements. 

“Formal specifications are hard because they need to describe complex software behavior precisely and concisely,” Zhai said. “They must be detailed enough to guide the AI, but not so detailed that they become another version of the code. This is especially hard when AI agents quickly change many connected parts of a project, because the specifications must keep up with the code, documentation, dependencies, and developer intent.” 

Zhai's research could reduce costly failures and improve the reliability of software used in high-stakes fields such as healthcare and finance, where safety and public trust are critical. 

Beyond improving safety and reliability, Zhai hopes the work will broaden access to software development. More dependable AI coding tools could help students, small businesses, public agencies, and community organizations turn ideas into functioning software without needing extensive technical expertise.  

“Many people have useful software ideas, but not the technical background to build reliable systems,” she said. “I hope this project can help AI turn those ideas into software people can actually trust.” 

The CAREER Award is one of the NSF’s most prestigious honors in support of early-career faculty who exemplify the role of teacher-scholars through outstanding research.