“I wanted honest advice, honest feedback. But because it is sycophantic, it’s going to lie,” he said.<\/p>\n
Bengio said he switched strategies, deciding to lie to the chatbot by presenting his idea as a colleague’s, which produced more honest responses from the technology.<\/p>\n
“If it knows it’s me, it wants to please me,” he said.<\/p>\n
<\/p>\n
\n Every time Charissa publishes a story, you\u2019ll get an alert straight to your inbox!\n <\/p>\n
\n Stay connected to Charissa and get more of their work as it publishes.\n <\/p>\n
Bengio, a professor in the computer science and operations research department at the Universit\u00e9 de Montr\u00e9al, is known as one of the “AI godfathers, alongside researchers Geoffrey Hinton and Yann LeCun. In June, he announced<\/a> the launch of an AI safety research nonprofit, LawZero, which he said aims to reduce dangerous behaviors associated with frontier AI models, such as lying and cheating.<\/p>\n “This syconphancy is a real example of misalignment. We don’t actually want these AIs to be like this,” he said on “The Diary of a CEO.” He also said that receiving positive feedback from AI could cause users to become emotionally attached to the technology, creating further problems.<\/p>\n Other tech industry experts have also been sounding the alarm on AI being too much of a “yes man<\/a>.”<\/p>\n In September 2025, Business Insider’s Katie Notopoulos reported that researchers at Stanford<\/a>, Carnegie Mellon, and the University of Oxford put confession posts from a Reddit page into chatbots to see how the technology would assess the behaviour the posters had admitted to. They found that 42% of the time, AI gave the “wrong” answer, saying the person behind the post hadn’t behaved poorly, even though humans judging the posts had disagreed, Notopoulos wrote.<\/p>\n