Apart from holding a dominated place among other competitors, industry leaders always contribute to the prosperity of the whole industry by making rules. The widely recognized rules can be of benefit to the well-placed enterprises and eject spoilers from the market.
As one of China’s artificial intelligence (AI) leaders, Megvii (旷视) published “AI Application Criteria (人工智能应用准则)” on July 8, which aims to provide guidance for the appropriate use of AI-related technologies and the sustainable development of the AI ecosystem.
The Criteria mainly consist of six dimensions: legitimacy, supervision of human, reliability and security of AI technology, fairness and variety, accountability and timely revision, data security and privacy protection. In each session, the criteria give specific manipulation standards, as well as response or solutions for every possible situation. For example, when it comes to the concept of “Supervision of Human”, the Criteria claim that AI should develop under the supervision of humans to protect the public from potential security risks.
By following this rule, Megvii hopes all the AI companies can maximize their business and society valuation by keeping self-critical. The company advocates thinking much of ethics and safety issues to avoid the misuse of technologies.
Founded in 2011, Megvii started as a facial recognition technology supplier based on its service platform Face++. Today, this company has grown to be an AI global leader with a highly competitive barrier in terms of deep learning, providing services to multiple scenarios.
According to Megvii, it has obtained more than 1,100 patents in fields of AI algorithms and applications, serving hundreds of thousands of AI developers and more than 1,000 industrial clients. Other than the first move in making rules, Megvii has participated in the process of conducting 19 AI-related criteria.
Megvii is not the first organization to conduct such industry rules. Last month, Chinese Management Professional Committee of New Generation AI (国家新一代人工智能治理专业委员会), an official management institute founded by Ministry of Science and Technology, issued a document named “New Generation AI Management Principle – Developing responsible AI”. This document introduces the framework and action directions of AI management, emphasizing eight principles including harmony and friendliness, fairness and justness, comprehension and available-of-sharing, respect of privacy, safeness and available-of-controlled, shared-responsibility, cooperativity, and flexibility of management. This document is regarded as a beginning for the further research of AI management, as well as a step forward for the construction of the AI management platform in China.
SenseTime (商汤科技), another AI leading company in China, had a similar attempt last year. The company signed a strategic cooperation agreement with China Electronics Standardization Institute (中国电子技术标准化研究院), a Chinese government research department. Under this agreement, SenseTime is responsible for establishing the standardization of computer vision technology.