MIT Technology Review answers reader questions about AI risks

MIT Technology Review published a Q&A based on a live subscribers’ event that explored whether artificial intelligence could cause serious harm, including through cyberattacks, biological threats, and autonomous agents. The piece, published on September 18, 2026, features responses from senior AI editor Will Douglas Heaven and AI reporter Grace Huckins. It discusses concerns about AI alignment, the difficulty of controlling autonomous systems, and why some researchers and companies are pushing for slower development while safety work continues. The article also examines whether fears about AI risk are being amplified for public relations purposes, and it argues that the more immediate challenge is reducing realworld harms from current systems while improving monitoring, transparency, and regulation.