What are the implications of AI accuracy suppression in software development?

A recent policy statement highlights concerns about the suppression of accuracy in AI systems. As software engineers and developers, how do you think this will impact our industry? Will it affect the trustworthiness of AI applications? What measures can we take to ensure our AI systems remain accurate and reliable? Share your thoughts and experiences.