Gates examines several immediate risks associated with artificial intelligence while arguing that they are manageable. He considers AI-generated misinformation and deepfakes, increasingly capable cyberattacks, disruption to employment, biased or inaccurate model outputs, and concerns about students relying on AI for written work. Rather than treating these problems as entirely unprecedented, he compares them with challenges created by earlier technologies, including automobiles, personal computers and the Internet.

Gates argues that responses should combine technical development, government action and social adaptation. AI itself may help detect manipulated media and security vulnerabilities, while laws will need to address areas such as deepfakes and fraud. Workers may require retraining, educators will need to adapt teaching practices, and AI companies should address privacy, bias and misuse. He distinguishes these current concerns from longer-term questions about superintelligent systems, concluding that societies should address present risks without abandoning the potential benefits of AI.