Artificial intelligence has promoted a series of innovations in social risk management, and it is becoming a hotly debated issue in academic circles at home and abroad. Through typical cases and theoretical research, it is found that in terms of “technologygovernance”, in the era of artificial intelligence, risk governance first needs to solve technical problems. Only by making the best choice of technology can the governance effect be optimized. As far as “governmentsociety” is concerned, artificial intelligence not only improves government governance capabilities, but also promotes civil society to become the main body of governance through complex adaptive systems, community participation, and community empowerment. In terms of “challengeprospect”, artificial intelligence will rapidly make human intelligence lag behind in many aspects. The most important technical problem of risk management will be taken care of by artificial intelligence. The phenomenon of socalled “technological singularity” or “singularity” will appear, which has bright application prospects. In terms of “deficiencytrap”, artificial intelligence lacking selfawareness is much more dangerous than the human brain. It may escape human control and create greater social risks. While managing social risks, it may also create new ones. Nevertheless, as a part of the society, artificial intelligence has not only strong social functions, but also other functions such as rapidity, accuracy, and versatility, and has promoted major changes in risk management. The governor should raise it to the strategic level and actively adapt to its development and changes.