Ethics

A new study by MIT researchers found that it is impossible to draw a direct link between AI diffusion model outputs and an artist’s body of work, writes Gizmodo’s Webb Wright. “If a [diffusion] model generates something, you want to be able to say, ‘Oh, this part of the training data was responsible,'” says Zheng Dai, PhD ’24, lead author on the study. “It’s important for us to understand how these models work to properly study them or regulate them.” 

Prof. David Gifford and Zheng Dai PhD ’24 discovered that AI diffusion models can replicate the style of an artist even when their work was omitted from datasets, leading to ‘attribution decay,’ writes Fast Company’s Jesus Diaz. “This trend, what the researchers call ‘attribution decay,’ means that artificial intelligence can produce an image that resembles a particular artist’s work, while having no provable causal link to that artist’s actual contribution to the training data,” explains Diaz.  

MIT researchers found the more data an AI model is trained on could complicate AI copyright, auditing, and governance, writes Computer World’s Taryn Plumb. “[S]ingle artworks by specific artists, or photographs of certain people, could be entirely removed from datasets, and the model could still reproduce that image or style,” writes Plumb. “Essentially, tangible connections are lost, and linking to specific data points responsible for generated samples is ‘practically impossible,’ or can even vanish, the researchers explained.” 

Prof. Emilio J. Castilla speaks to Forbes’ Lauren Russell about how bias can shape employee performance reviews. “Bias—whether explicit or unconscious—can potentially shape employment decisions at every stage, from who is interviewed and hired to who is promoted and who is rewarded for excellent performance,” says Castilla.  

Chronicle of Philanthropy reporter Maria Di Mento spotlights how the creation of the MIT Schwarzman College of Computing allowed MIT to develop new “interdisciplinary programs to prepare students for an AI-saturated world and help them understand the social and ethical implications of digital technologies.” Prof. Daniel Huttenlocher, dean of the Schwarzman College of Computing, explains that: “MIT realized that effective education in the age of AI has to look different than it has in the past. Traditional siloing of expertise won’t work when AI is expected to touch nearly every part of people’s lives and is changing the way people in disciplines outside of computing are advancing their work.”

Writing for Forbes, Senior Lecturer Guadalupe Hayes-Mota 08, SM '16, MBA '16 explains the CARES framework, a new model for founders looking to integrate ethics into their company.  “Biotechnology can cure, extend and transform life,” writes Hayes-Mota. “But based on my experiences, its full potential can only be realized when innovation and ethics advance together. To every scientist-founder at the edge of discovery: How can your business plan not only explain how you’ll succeed, but also why your success should exist?” 

Writing for Forbes, Senior Lecturer Guadalupe Hayes-Mota '08, SM '16, MBA '16 emphasizes the importance of implementing ethical frameworks when developing AI systems designed for use in healthcare. “The future of AI in healthcare not only needs to be intelligent,” writes Hayes-Mota. “It needs to be trusted. And in healthcare, trust is the ultimate competitive edge.” 

Financial Times reporter Melissa Heikkilä spotlights how MIT researchers have uncovered evidence that increased use of AI tools by medical professionals risks “leading to worse health outcomes for women and ethnic minorities.” One study found that numerous AI models “recommended a much lower level of care for female patients,” writes Heikkilä. “A separate study by the MIT team showed that OpenAI’s GPT-4 and other models also displayed answers that had less compassion towards Black and Asian people seeking support for mental health problems.” 

Prof. Marzyeh Ghassemi speaks with Boston Globe reporter Hiawatha Bray about her work uncovering issues with bias and trustworthiness in medical AI systems. “I love developing AI systems,” says Ghassemi. “I’m a professor at MIT for a reason. But it’s clear to me that naive deployments of these systems, that do not recognize the baggage that human data comes with, will lead to harm.”

Writing for Fast Company, Senior Lecturer Guadalupe Hayes-Mota SB '08, MS '16, MBA '16, explores new approaches to improve the drug development process and more effectively connect scientific discoveries and treatment. “Transforming scientific discoveries into better treatments is a complex challenge, but it is also an opportunity to rethink our approach to healthcare innovation,” writes Hayes-Mota. “Through cross-disciplinary collaboration, leveraging AI, focusing on patient-centered innovation, and rethinking R&D, we can create a future where scientific breakthroughs translate into meaningful, accessible treatments for all.”

Researchers at MIT have found that commercially available AI models, “were more likely to recommend calling police when shown Ring videos captured in minority communities,” reports Kyle Wiggers for TechCrunch. “The study also found that, when analyzing footage from majority-white neighborhoods, the models were less likely to describe scenes using terms like ‘casing the property’ or ‘burglary tools,’” writes Wiggers. 

Prof. Devavrat Shah is interviewed by Forbes’ Gary Drenik on balancing AI innovation with ethical considerations, noting governance helps ensure the benefits of AI are fairly distributed across society. “Our responsibility is to harness [AI’s] potential while safeguarding against its risks,” Shah explains. “This approach to promoting responsible AI development hinges on governance rooted in collaboration, transparency and actionable guidance."

Prof. David Rand speaks with VOA News about the potential impact of adding watermarks to AI generated materials. “My concern is if you label as AI-generated, everything that’s AI-generated regardless of whether it’s misleading or not, people essentially are going to stop really paying attention to it,” says Rand.