By Grace Stanley
Artificial intelligence is everywhere, from the apps people use to the systems that shape hiring decisions and healthcare. But what happens when these tools don’t work equally well for everyone?
That question drives the research of Allison Koenecke, a new assistant professor of information science at Cornell Tech and the Cornell Ann S. Bowers College of Computing and Information Science. Koenecke’s work has helped expose disparities in speech-to-text technologies, which can misinterpret speakers with minoritized dialects or health conditions — leading to real-world consequences in medicine, employment, and beyond.
Koenecke’s research has been cited in legal cases, including one brought by the American Civil Liberties Union (ACLU) against a hiring platform that failed to accommodate a Deaf applicant. Additionally, her work has been cited in investigative reporting, including a piece in AP News about the potential for hallucinations in medical transcriptions.
Honored as a Sloan Fellow and a Forbes 30 Under 30 recipient in science, Koenecke has focused her research on algorithmic fairness. Her work has been featured in prominent outlets such as The New York Times, The Atlantic, and Scientific American. She has also been published in top academic conferences, including PNAS, NeurIPS, and FAccT. She is the recipient of several NSF grants and a Faculty of the Year Award in Diversity, Equity, Inclusion, and Belonging from Cornell Bowers.
After teaching data science to hundreds of undergraduates in Ithaca, Koenecke is excited to bring her expertise to Cornell Tech in New York City — and to collaborate with local organizations working on speech systems and equity-focused tech.
What is your academic and research focus?
Much of my work focuses on algorithmic fairness, which can have different meanings to different people. For me, it includes studying how AI systems affect marginalized populations and how we can improve public services — such as healthcare or government systems — to serve everyone better. I’ve worked extensively on speech-to-text algorithms and large language models, looking at how they perform across different demographics. For example, these models often perform poorly for people with minoritized dialects or speech impairments, which can lead to inaccurate transcriptions and downstream harm.
What motivated you to come to Cornell Tech?
I’m super excited about the on-the-ground collaborations I can build in New York City. Being here allows me to work more directly with impacted populations and partner with organizations like Weill Cornell Medicine, which is doing great work on speech systems. I’m also looking forward to working with my colleagues at Cornell Tech and contributing to the broader tech ecosystem in the city.
What societal impacts are you most concerned about in your research?
Speech is a modality that people don’t always realize is used in so many basic applications. It’s used to transcribe patient notes in doctors’ offices and applicant interviews in hiring pipelines, such as those run by platforms like HireVue. Errors in speech recognition can lead to worse medical diagnoses or biased hiring outcomes. One example I’m proud of is a case brought by the ACLU against HireVue for failing to accommodate a Deaf applicant. They cited research that my students and I conducted, which was incredibly meaningful. It’s powerful to see how data and research can support legal efforts to fight discrimination.
What courses are you teaching or planning to teach at Cornell Tech?
For the past three years, I’ve taught “Introduction to Data Science” to over 200 undergrads each fall semester in Ithaca. It’s been a great way to understand what excites students about data science and how they think about its applications. I’m now bringing that same energy to Cornell Tech, where I’ll be teaching other data science-related courses, including one titled “Data Science in the Wild.” Stay tuned!
What do you like to do outside of research and teaching?
Outside of work, I’m excited to explore all the great food New York City has to offer. It’s one of the things I’m most looking forward to about being here!
Grace Stanley is the staff writer-editor for Cornell Tech.
