Artificial Intelligence: Who Is Responsible When AI Goes Wrong?

Are we in a love-hate relationship with Artificial Intelligence? Some people have been using ChatGPT to fix code at work, find relationship therapies, seek advice during moral dilemmas, and even save their pets’ lives. 

While others have argued about AI’s ethical and legal problems, issues related to privacy and legitimacy, including algorithmic transparency, cybersecurity vulnerabilities, discrimination, and lack of accountability, have been brought into the public view.   

More importantly, who’s responsible when a self-driving car gets into an accident? What happens if ChatGPT leaks copyrighted or protected information? 

Dr Sophie Calabretto talks to Cosmos Magazine journalist Petra Stock about how our legal systems are not ready for an AI future and the exploited labour behind artificial intelligence. 

Stock says most government policies and legal systems struggle to keep up with AI, while some groups and organisations, such as Algorithmic Justice League (AJL), have stepped forward and called for fairer and more inclusive algorithms.  

“The core goal of AJL is to raise awareness about issues such as legal danger and privacy. Their goal is also to work with policymakers to prevent harm,” Stock said. 

As part of the process to tackle the legal problems of AI, one of the founders of AJL represented the league at a congressional hearing in 2019 to discuss the facial recognition technology being used commercially and by the government.  

“The founder spoke about the shortcomings of facial recognition in identifying people with dark skin and also feminine features,” Stock explained. 

Tune into the full episode of Science Explained with host, Dr Sophie Calabretto to hear the full recap of this month’s discoveries. 

Introducing The Science Briefing: a podcast about the science of everything and your new go-to podcast for your snapshot of science news. Hosted by Dr Sophie Calabretto and featuring journalists from Cosmos Magazine.