April 24, 2024

UAntwerp focuses on ethical aspects of AI

With the foundation of the new Antwerp Center on Responsible AI (ACRAI), UAntwerp underlines that ethical decisions relating to artificial intelligence are too important to leave to companies alone. The new centre brings together academic expertise from six faculties, in order to promote responsible use of AI.

Artificial intelligence, fed by huge amounts of information, dishes up the “optimal” answer to our query in no time. Useful? Sure! But not without risk, for one thing because an algorithm is susceptible to prejudice, which means the dangers of discrimination and racism are always lurking in the background when it comes to AI-driven decisions.

‘Take Geoffrey Hinton, one of the founding fathers of AI technology. In the aftermath of discussions revolving around generative AI, such as ChatGPT, he left Google and warned us about the dangers of the technology,’ says Professor David Martens. ‘However, ChatGPT didn’t introduce these risks: as part of the Data Science Ethics course, I’ve spent years teaching about the risks of manipulation, discrimination, the lack of transparency in AI’s data sources and decision making, privacy issues, etcetera.’

Racist chatbots

‘Incidents we’ve seen include Microsoft’s racist chatbots of a few years ago, sexism in Amazon’s AI-driven recruitment tool, and Netflix’s disclosure of a dataset revealing sexual orientation. What’s becoming increasingly clear is that the big tech players, such as OpenAI, Microsoft, Meta and Google, lead the discussion on this theme. OpenAI’s CEO, for example, first asked the American Senate to regulate AI, and then threatened to withdraw OpenAI from Europe if the continent were to “overregulate” AI. We need academic voices and scientific research to counterbalance all of this.’

These kinds of challenges demand a multidisciplinary approach. This is why the University of Antwerp’s ample expertise in various domains, including computer science, law, philosophy and sociology, is now being brought together in the Antwerp Center on Artificial Intelligence (ACRAI). Through ACRAI, UAntwerp is hoping to raise awareness of the problems and contribute to solving them.

Check the ACRAI website here: www.uantwerpen.be/responsibleai

Related News