Workflow
智能财务迎关键性制度升级,大模型应用潜力与挑战并存
2 1 Shi Ji Jing Ji Bao Dao·2025-07-29 09:49

Core Insights - The year 2025 is identified as a pivotal point for the digital transformation of the finance industry, driven by policies and advancements in AI technology [1][2] Policy and Regulatory Developments - A series of policies and regulations are being introduced to facilitate the implementation of intelligent finance, including the incorporation of "accounting informationization" into the revised Accounting Law of the People's Republic of China [3] - New standards and regulations, such as the Accounting Informationization Work Specification and the Basic Functions and Service Standards for Accounting Software, are being enforced to ensure compliance and enhance data security [3][4] - The promotion of electronic vouchers and the push for comprehensive electronic invoicing by 2025 signify a shift towards big data in tax administration [3] AI Integration in Finance - A significant percentage (69.03%) of finance professionals believe that AI can be extensively utilized in the accounting sector, up from 64.73% in 2024, indicating growing confidence in AI's role [2] - The integration of AI is transforming accounting functions from value reflection and accounting to value creation, enhancing management decision-making and accelerating the integration of finance and business [2][5] Industry Applications and Innovations - Financial institutions, such as Shanghai Pudong Development Bank, are leveraging AI for marketing and credit analysis, showcasing a new model of banking services [6] - The payment industry is also evolving, with companies like Huifu Technology introducing intelligent payment solutions that can generate customized payment plans in real-time [6] Concerns and Challenges - Despite the advancements, there are concerns regarding the risks associated with emerging AI technologies, particularly in areas like data security and compliance [7] - The industry remains cautious about the application of large models in finance, citing issues with consistency and a lack of real-world case studies to support their use [7]