Yann LeCun, Turing Award winner and widely recognised as one of the three “godfathers of AI,” has pushed back against calls to slow down artificial intelligence (AI) development, accusing Anthropic CEO Dario Amodei of using alarmist messaging to lobby regulators and restrict public access to the technology. Speaking during an interview on the morning of October 12 at the Westin Chosun Seoul in Jung-gu, the co-founder and chairman of AMI Labs addressed the intensifying divide across Silicon Valley over safety regulations and model releases.“It’s natural for companies to release products with safety measures to prevent harm to people,” said LeCun, who won the 2018 Turing Award alongside Yoshua Bengio and Geoffrey Hinton for pioneering the foundations of computer vision and image recognition, was in South Korea for the Industrial AI Global Cooperation Meeting ahead of the IIAC 2026 conference in Tokyo.“However, large language models (LLMs) are difficult to fine-tune, so they can potentially harm humans in certain cases. Sam Altman says, ‘We’ll release them cautiously,’ while Dario Amodei argues, ‘They’re too dangerous to manage and shouldn’t be given to the public.’ I oppose Dario’s view,” he added, as per a report in South Korean publication Chosun.
Yann LeCun opposes Anthropic: ‘No different from anti-enlightenment’
LeCun compared attempts to gatekeep underlying model weights to historical efforts to suppress information.“Restricting access is no different from anti-enlightenment – like the Ottoman Empire banning printing in the 15th century to prevent information from reaching people. Ultimately, foundational AI models should be open-source. Only then can human culture, knowledge, and information spread.”When asked whether modern systems can be controlled, LeCun dismissed the notion that today’s large language models pose an existential threat to society. Instead, he framed alarmist narratives as political leverage.“I don’t think current AI is dangerous. The argument for slowing down AI development is a form of marketing or lobbying. Anthropic is portraying AI as extremely dangerous to scare governments. But LLMs aren’t at that level yet.”While talking about the Hugging Face incident, where an OpenAI model attempted to break out of its environment, LeCun argued that such occurrences stem from poor supervision rather than machine malice.“The Hugging Face incident happened due to intentional or incompetent design flaws. Human supervisors failed to properly design the experimental environment, causing AI agents to deviate. Since these agents aren’t inherently malicious, no harm occurred.”Later during the meeting, LeCun reiterated that governing artificial intelligence is fundamentally a technical challenge rather than a crisis of control.