[2] The bulk has been about generative AI models, particularly large language models. Do the guardrails take into account that AI models are hard to explain, from time-to-time hallucinate, and may strategically deceive users? In essence, how does one hold liable the persons who deploy AI models that create truly unpredictable harm? In the real world, AI models themselves also can hallucinate but don’t necessarily have Morpheus there to save them. [28]If an AI model can hallucinate a bad case citation, couldn’t an AI model used by a broker or investment adviser hallucinate an unsuitable or conflicted investment recommendation?

Source: New York Times February 13, 2024 17:40 UTC