In the context of rapid advancements in information technology, social media platforms such as Weibo, WeChat, and Douyin have become crucial channels for accessing information and understanding current affairs. Despite the convenience these platforms offer, the issue of false information has become increasingly severe due to the lack of effective regulatory mechanisms and rapid response measures. Traditional manual methods for detecting false information are inefficient and resource-intensive, necessitating accelerated digital development to enhance digital societal infrastructure and improve the level of digital and intelligent public services and social governance. To address this, this paper proposes a false information identification method that combines BERT and ResNet models for efficient and automatic detection of false information on social media. This approach aims to enhance the efficiency of false information detection, curb its spread, and promote the construction of a digital society.