Recently, the 2022 Trustworthy AI Summit (main forum) hosted by the China Academy of Information and Communication Research, the National Engineering Research Center for Deep Learning Technology and Applications, and the Key Laboratory of Artificial Intelligence Key Technology and Application Evaluation of the Ministry of Industry and Information Technology was successfully held.
At the meeting, the member list of the first batch of trusted artificial intelligence promotion plan (referred to as "Zhixin" plan) launched by the Cloud Computing and Big Data Research Institute of the Chinese Academy of Information and Communications was announced, and Xiaoshi Technology was successfully selected as one of the first member units.
The "Smart Trust" plan promotes the improvement of AI's credible governance capabilities
With the development of the new generation of information technology, the depth and breadth of artificial intelligence applications are constantly expanding, and have become an important component of information infrastructure. The information security risks caused by artificial intelligence also come with it, and "trust anxiety" permeates the world. The international community's demand for trusted artificial intelligence is becoming increasingly urgent.
The original intention of the "Zhixin" plan launched by the Cloud Computing and Big Data Research Institute of the China Academy of Information and Communications is to promote the improvement of artificial intelligence's trustworthy governance capabilities, accelerate the reaching of industry consensus, and ensure the rapid, healthy, and sustainable implementation of intelligent applications.
This plan is intended to target artificial intelligence related entities such as government, industry, academia, research, and application. It mainly carries out technical research, standard formulation, testing and evaluation, supply and demand coordination, industrial exchange, event summits, platform construction, case promotion, and other work. The main work includes conducting research on trusted artificial intelligence technology and industry, developing artificial intelligence standards for trusted implementation, and building a trusted artificial intelligence industry ecosystem.
As a representative enterprise in the field of artificial intelligence in Jiangsu Province, Minivision Technology has always been highly concerned about the development of trustworthy artificial intelligence. This successful selection fully affirms the practical value released by Xiaoshi Technology in promoting the implementation of trustworthy AI applications, and also demonstrates Minivision Technology's technical strength in assisting industry development.
AI landing, trust first
In practice, Minivision Technology has gradually formed a trustworthy system of team management, data management, technology research and development, and standards leading the "four-wheel drive", continuously promoting the development of trustworthy artificial intelligence.
In terms of team management, Minivision Technology has established a multi background team in research and development, legal and other fields to ensure the safety and credibility of the entire process of technology product development and application, and continuously improve personnel's "credibility" awareness and technical expertise.
In terms of data management, develop and implement data application management mechanisms, establish approval and review processes and project differentiation authorization mechanisms, deploy operation and maintenance audit systems, establish dedicated computer room management data servers, and establish strict access management systems.
In terms of technological research and development, Minivision Technology fully embeds the requirements of trusted artificial intelligence throughout the entire process. At the same time, a trusted AI governance framework is constructed from aspects such as trusted data, trusted algorithms, trusted computing, and trusted management. Components such as anomaly data detection and bias elimination are utilized to process abnormal and heterogeneous data, enhancing the robustness of the model; Add algorithm components such as causal algorithms and fairness algorithms to enhance the model's adaptability and defense capabilities.
In terms of standard leadership, Minivision Technology became a member unit of the "Trusted Face Recognition Guardianship Program" of the China Academy of Information and Communications Technology in 2021 and a member unit of the Trusted Working Group of the Artificial Intelligence Subcommittee of the National Information and Communications Commission. At the same time, actively participate in the development of multiple industry, group, and landmark landmarks, emphasizing the importance of protecting residents' privacy from the perspectives of engineering design, data application, etc.
Trusted artificial intelligence is a long-term and systematic project that requires joint efforts from multiple parties to achieve.
In the future, Minivision Technology will also fully fulfill its "trustworthiness" obligation, carry out research projects on trusted artificial intelligence technology and applications, build a trusted artificial intelligence standard system around research and development, technology, products, applications, and management, utilize the main forces of "government, industry, academia, research, and application" artificial intelligence, accelerate the formulation of trusted artificial intelligence standards, and promote the high-quality development of trusted artificial intelligence.
