Recently, the management regulations on data development have continued to be a focus of attention from all parties, as the list of pilot consortia for the construction of high-quality industry datasets empowered by artificial intelligence has been announced, and behaviors such as big data killing have been clearly prohibited. Data is an important engine driving innovation, but it also harbors security, governance, and legal challenges. It is crucial to balance development and security, stimulate innovation vitality, prevent major risks, and continuously improve data legislation.
China continues to make efforts in data legislation, and has initially established a legal regulatory system based on the Cybersecurity Law, Data Security Law, Personal Information Protection Law, etc. At the level of data elements, a series of laws and regulations provide a hierarchical and categorized compliance path for data circulation. At the level of algorithms and models, a regulatory baseline centered on "filing identification traceability" has been formed for different technological forms such as recommendation algorithms, deep synthesis, and generative AI. In addition to laws and regulations, national standards, industry guidelines, and other tools are also used, such as the "Basic Requirements for Security of Generative Artificial Intelligence Services", which provide specific and actionable guidance for enterprises.
However, some data security risks still exist at present. For example, artificial intelligence is deeply integrated with various fields, but its dependence on massive amounts of data has greatly amplified risks such as data leakage, illegal acquisition, and malicious "poisoning", directly threatening personal privacy, trade secrets, and even national security; The "black box" and bias of algorithmic decision-making challenge traditional principles of transparency and fairness, implying imperceptible discrimination, and the opacity of the decision-making process can also pose challenges to individual rights protection and accountability mechanisms. In addition, the use of training data often conflicts with personal information protection and intellectual property rights (such as copyright). In view of this, it is necessary to clarify the rights and responsibilities of all parties under the premise of legality and compliance, establish a predictable and traceable governance framework, and build a scientific, systematic, dynamic, and forward-looking data legal system to empower technology and provide a "security lock" for people's beautiful "digital life".
On the one hand, we adhere to the principle of inclusiveness and prudence. Relevant legislation needs to seek a balance between the stability of the rule of law and technological flexibility, and innovation cannot be stifled by the pursuit of "absolute security". A differentiated regulatory framework based on risk levels should be established, providing loose development space for low-risk applications and implementing strict regulations for high-risk areas such as healthcare and finance. Accelerate the improvement of data property rights and circulation system. Explore the structural separation system of data property rights, distinguish the rights of data sources, holders, and processors, and balance the interests of all parties. Refine the rules for data rights confirmation, authorization, transactions, and profit distribution, while ensuring national security and personal privacy. Through standard contracts, security authentication, and other tools, connect cross-border and domestic data circulation channels, and solve the dilemma of "data shortage" and "compliance difficulties" faced by artificial intelligence enterprises. Strengthen algorithmic accountability and artificial intelligence infringement adjudication rules. Regarding the "algorithmic black box", especially in decision-making involving significant personal rights or public interests, algorithmic transparency should be gradually promoted, and infringement judgment rules should be improved from a judicial perspective to make clearer provisions on the allocation of liability for damages caused by AI, as well as remedies for bias and discrimination.
On the other hand, building a multi-dimensional and collaborative normative system of "law standard ethics". Through diverse governance, a governance force covering the entire lifecycle of technology is formed, especially by embedding ethical standards into the entire process of technology research and application, ensuring that the development of artificial intelligence aligns with human values. Leading the deepening of global artificial intelligence governance, actively responding to the common concerns of the world about the safety, trustworthiness, and controllability of artificial intelligence, promoting mutual recognition of international rules in cross-border data flow, ethical standards, security assessments, and other aspects through bilateral and multilateral dialogue, and contributing Chinese wisdom and solutions to building a fair, reasonable, and inclusive global artificial intelligence governance system. (Source: Economic Daily, Author: Li Yi)
(Editor in charge: Nian Wei)
