Skip to main content
Markkula Center for Applied Ethics Homepage

Stories

Forbes Logo: A white capital F on black background.

Forbes Logo: A white capital F on black background.

Four Things To Settle Before You Release A Biology AI Model

Guadalupe Hayes-Mota, published by Forbes Magazine.

Director of Bioethics, Guadalupe Hayes-Mota, writes about biotech AI models that could potentially serve as a tool that helps a therapeutic delivery virus evade immune defenses, but could also help malicious actors synthesize harmful pathogens.

Hayes-Mota says the practical decision for these technologies still lands on company leaders, and His CARES model–an ethical AI framework for healthcare–won't replace the need for established safety practices, but can help leaders organize their decisions around five ethical responsibilities that can help make accountability more explicit.

"Responsible governance should be affordable enough to implement and credible enough to withstand scrutiny," says Hayes-Mota

This is my proposed governance tool, not an established industry standard, but the idea draws on Magnifica Humanitas, Pope Leo XIV’s encyclical on AI. It describes the common good as “the social expression of the dignity recognized in every person.” The key is every person, not every paying customer.

For a capsid-design model, two populations are missing from the commercial agreement. One bears the risk: people potentially harmed by misuse. The other misses the benefit: patients and researchers excluded by cost, geography or institutional status.

 

Guadalupe Hayes-Mota, published by Forbes.

 

Ethics
media, bioethics