AI Safety Concerns Grow as Machine-Learning Pioneer Warns of Malicious Use

Yoshua Bengio, a renowned machine-learning pioneer, has expressed concerns about the malicious use of artificial intelligence, stating that such use is already happening. Bengio, considered one of the founders of AI, has long been at the forefront of machine-learning research. However, his opinions on the technology have shifted in recent years as he has come to realize the potential risks posed by AI. In an interview with Nature, Bengio emphasized the need for developing AI with safety built in from the start. He believes that the current approach to AI development, which focuses on creating more powerful and efficient models, is not sufficient to address the risks associated with AI. Bengio's comments come as the AI community grapples with the potential consequences of creating autonomous systems that can learn and adapt at an exponential rate. According to Bengio, the malicious use of AI is already happening, and it is essential to take proactive measures to prevent such use in the future. He cited the example of AI-powered deepfakes, which can be used to create realistic but fake videos and audio recordings. Bengio warned that the spread of such technology could have severe consequences, including the manipulation of public opinion and the spread of misinformation. Bengio is not alone in his concerns about the risks associated with AI. Many experts in the field have expressed similar concerns, highlighting the need for more research into the development of safe and responsible AI systems. The AI community has been working on developing techniques to detect and prevent malicious use of AI, but more work is needed to ensure that AI systems are designed with safety and security in mind. In recent years, Bengio has been working on developing an AI system that is designed with safety and security in mind. He has been exploring the use of techniques such as value alignment, which aims to align the goals and values of AI systems with those of humans. Bengio believes that this approach can help to prevent the malicious use of AI and ensure that AI systems are developed in a responsible and sustainable way. The development of safe and responsible AI systems is a complex task that requires the collaboration of experts from various fields, including computer science, philosophy, and ethics. Bengio's work on AI safety is part of a broader effort to develop AI systems that are designed to benefit society as a whole. As the AI community continues to grapple with the potential risks and benefits of AI, Bengio's work serves as a reminder of the need for responsible and sustainable AI development. In an email to Nature, Bengio emphasized the importance of developing AI systems that are transparent, explainable, and accountable. He believes that these characteristics are essential for building trust in AI systems and ensuring that they are developed in a responsible and sustainable way. Bengio's work on AI safety is ongoing, and he continues to collaborate with experts from around the world to develop more effective techniques for preventing the malicious use of AI. The Nature Podcast Extra, released on November 14, 2025, features an interview with Bengio, in which he discusses his concerns about the risks associated with AI and his efforts to develop safe and responsible AI systems. The podcast is available on Apple Podcasts, Spotify, YouTube Music, and other popular podcast platforms. As the AI community continues to develop and deploy more advanced AI systems, the need for safe and responsible AI development becomes increasingly pressing. Bengio's work serves as a reminder of the importance of prioritizing AI safety and security, and his efforts to develop more effective techniques for preventing the malicious use of AI are an essential part of this effort.