【摘要】 计算宣传作为AI时代算法技术与政治宣传深度耦合的产物，正在深刻重塑当代国际宣传和传播方式。随着智能化舆论博弈的加剧，美西方试图利用智能技术在信息生产与传播方面的先发优势，操纵政治宣传，抹黑中国政治形象。尤其在美西方对华算法政治操弄下，中国面临网络空间治理难度加剧、社会群体认知体系失调、民众政治制度信任度降低、主流意识形态传播引领效能减弱等一系列政治安全风险。为了有效应对美西方对华计算宣传，在技术防御层面，应强化算法技术保障与提高计算宣传风险预警能力；在制度设计层面，构建一核多元的网络空间治理格局；在议题设置层面，构建攻防兼备的对外传播体系；在素养培养层面，推动全民信息能力与数字素养提升工程；在国际合作方面，携手各国构建计算宣传防御网络。更多还原

【Abstract】 As a product of the deep integration between algorithmic technology and political propaganda in the AI era, computational propaganda is profoundly reshaping contemporary international propaganda and communication methods. With the intensification of intelligent public opinion competition, Western countries are attempting to leverage their firstmover advantages in the intelligent production and information dissemination to manipulate political propaganda and tarnish China’s political image. Particularly under the algorithmic p olitical manipulation employed by Western nations, China faces a series of political security risks, including increased difficulty in cyberspace governance, disruption in the cognitive systems of social groups, declining public trust in political institutions, and diminished effectiveness in disseminating and guiding mainstream ideology. To effectively counter Western computational propaganda against China, efforts must be made across multiple dimensions: at the technical defense level, algorithmic safeguards should be strengthened, and the capability to detect risks related to computational propaganda should be enhanced; at the institutional design level, a centralized and diversified governance framework for cyberspace should be established; at the agenda-setting level, a comprehensive external communication system that balances offense and defense should be constructed; at the capacity-building level, initiatives to improve public information literacy and digital skills should be promoted; and in terms of international cooperation, collaborative efforts should be made with other countries to build a defense network against computational propaganda.更多还原