A few weeks ago, Amazon and Hopkins revealed the JHU + Amazon Initiative for Interactive AI (AI2AI), a research-based collaboration to further advance artificial intelligence (AI) research in areas such as machine learning, computer vision and speech processing.
Amazon, however, has a well-documented history of unfair and inhumane labor practices. The company has been accused of mistreating its employees, neglecting their safety and interfering with their efforts to unionize. Given this, we question their partnership.
Just like Amazon, Hopkins is also guilty of controversial and negligent labor practices. In 2019, the National Labor Relations Board found that Hopkins Hospital attempted to impede nurses’ unionization. In 2020, Teachers and Researchers United (TRU), an unofficial graduate student union at Hopkins, held a rally to demand that the University improve graduate students’ safety and security. The University has also ignored TRU’s demands to be recognized as an official union.
In addition to not supporting their employees, both Amazon and Hopkins do not pay their fair share of taxes. Amazon’s annual financial report, released in February, reveals that the company has avoided roughly $5.2 billion in corporate federal income taxes in 2021, despite its record earnings. The University also avoids paying property taxes due to its nonprofit status.
There are also widespread concerns over Amazon's improper data sharing and privacy violations. For example, Amazon employees listen to Alexa speaker recordings to improve human speech accuracy, often without users’ knowledge.
Is a partnership with Amazon really what we want? How will this partnership look under the scrutiny of the public eye? Even within the University, concerns about mentioning Amazon by name in the initiative’s title were brought to the Office of the President. Clearly, University administrators have considered the potential for controversy surrounding its association with the Amazon brand.
Even the founding director of AI2AI acknowledged the risk of controversy in an interview with The News-Letter. According to him, the University is not opposed to ending the partnership if something contentious occurs. Where is Hopkins drawing the line? We want more clarity over the circumstances that would cause the University to sever the relationship.
Through this partnership, Amazon will provide funding to the University for fellowships and research projects. We know research funding is important to Hopkins; for over 40 years, Hopkins has had the highest research and development spending of any U.S. university. We acknowledge that students benefit from the research opportunities here, but we hope our institution will not compromise its ethics with this partnership.
AI research in particular presents more challenging ethical questions, making it a hot button issue. For example, facial recognition software used by police is reported to replicate harmful biases and exacerbate discrimination. In fact, Amazon’s own AI recruiting engine was scrapped because it was biased against women.
We recognize the potential benefits of AI research; there are plenty of AI projects at Hopkins with commendable objectives. Take, for example, Professor Alan Yuille, who is working on deep learning algorithms to detect pancreatic cancer in CT scans. And last semester, the University received a $20 million grant to develop AI that promotes healthy aging.
As AI2AI progresses, we hope that Hopkins continues to focus on research into ethical AI. Namely, we would like to see the initiative prioritize eliminating discriminatory biases and protecting user privacy.
As students, we do not decide who Hopkins collaborates with. The University, however, must keep students informed when pursuing initiatives with such controversial partners.