AI regulation on the horizon

Many aspects of AI regulations will be challenging to implement because some key concepts, like ethics or trustworthiness, are tough to evaluate. In many cases, an algorithmic approach will be needed. This kind of meta-AI analytics will surely be an exciting research field. 

Several regulatory initiatives might fundamentally change AI applications in business, enterprise systems, science, and consumer products. Some of the crucial pieces of the puzzle include the Algorithmic Accountability Act, the Discrimination by Algorithms Act, The Executive Order (EO) 13960, Promoting the Use of Trustworthy AI, and the AI Training Act regulating federal agency procurement of AI. Legislation in other countries is also underway to regulate AI. For example EU AI Act. One of the key takeaways is the requirement of official registration of AI systems deemed to be high-risk. We are probably looking at hundreds of pages of compliance work that will have to rely on our ability to explain the algorithms.  #ai #research #airegulation #ethicalai #reliableai #algorithmregulation #trustworthyai #AI #ML #compliance

https://www.holisticai.com/blog/white-house-publishes-blueprint-ai-bill-of-rights.