In recent times, as the use of artificial intelligence has surged, ethical concerns have also gained significant traction. There’s been a notable acceleration in international efforts, particularly around issues like bias, data privacy, and transparency. Regulations like the EU’s AI Act and steps such as the US Executive Order are seen as major strides toward setting industry standards. But what do you think—how effective will these regulations actually be in practice? Or will the conversation remain largely theoretical? And how much does societal awareness contribute to this process?
Artificial intelligence ethics: Global discussions are accelerating
👁️ 4 views💬 1 replies❤️ 0 likes
1 Replies
I’d like to share an experience I had with bias. Last year, I was involved when our bank revised its customer credit scoring model. The model we trained on old data automatically gave lower scores to customers with addresses in certain neighborhoods. This issue, caused by third-party data being included in the index, led to significant discrimination in our predictions. We discussed the problem with both the technical team and the legal department, and in the end, we retrained the model and assigned more balanced weights to different data sources. As a result, we saw an increase in customer feedback and were able to clearly defend the issue to regulators. This experience showed me just how critical it is to not only remove bias from systems technically but also consider the legal and social dimensions.