Everyone agrees that artificial intelligence has the potential to dramatically improve people’s lives. But what if, intentionally or not, it also reinforces their biases?
That’s the concern raised by a paper published in September 2019 by the Berlin think-tank Stiftung Neue Verantwortung (New Responsibility Foundation). The essay is titled “Artificial Intelligence Needs Human Rights.” It says that the debate on how to make AI “ethical” has ignored other problems such the replication of cultural prejudices in algorithms.
Citing the dominance of American and Chinese developers of AI, the paper’s authors argue: “We must stop focusing the discourse on ethical principles, and instead shift the debate to human rights.”
They add: “If AI technologies were subject to scrutiny on the international stage as violations of human rights, rather than just harmful and unethical, then there would be more enforcement mechanisms with which to regulate the key innovators and perpetrators.”