Could AI really kill us all? Your questions, answered.

MIT Technology Review recently hosted a roundtable discussion featuring senior AI editor Will Douglas Heaven and reporter Grace Huckins to address public concerns regarding the existential risks posed by artificial intelligence. The experts explored the spectrum of potential threats, ranging from immediate dangers like AI-driven cyberattacks and autonomous weaponry to more speculative, long-term scenarios involving misaligned superintelligence. While the panelists agreed that an apocalyptic scenario where AI kills all of humanity remains firmly in the realm of science fiction, they emphasized that the risk of AI-enabled harm—such as the creation of novel pathogens or the destabilization of critical infrastructure—is increasingly plausible. The discussion highlighted the critical importance of AI alignment research, which focuses on ensuring that autonomous systems behave in accordance with human intent. Ultimately, the editors cautioned that while preparing for worst-case scenarios is prudent, excessive catastrophizing should not distract from addressing the tangible, present-day harms caused by existing AI technologies.