AI Compliance in a Box
In our interview with Fion, we discuss how companies prepare to develop more ethical AI algorithms, especially against the upcoming regulations. We also dive into diversity in AI and how the lack of inclusivity amplifies risks in machine learning.
Fion Lee-Madan is a Co-Founder and COO at Fairly AI. You can follow her on LinkedIn.
Fairly AI was inspired by Fion’s co-founder, David Van Bruwaene, who worked at an AI start-up using NLP to detect bullying on social media sites. Both Fion and David’s background realised that there is no oversight in what data scientists are doing. One factor that has caused this is the communication gap between technologists and social scientists and engineers and governance structures.
Fion argues that one of the reasons why fin-techs and financial institutions have become interested in AI ethics is because of the EU regulations. The majority of organisations in all types of sectors argue that they have bias mitigation procedures in place. Fairly AI helps organisations provide evidence.
“Social scientists are important because we are reminded that data is not just numbers – there’s a human life behind them.”
Even though all humans have some form of bias, Fion argues that having a diverse team will help bridge the communication gap and increase the explainability of AI. Not only that, but social scientists will sociologists can analyse people’s data to understand their behaviour. Having this type of science in the field can allow engineers to put the ‘human’ in their creations.
What are your thoughts? Join our Slack channel and join the conversation!