Google has laid out a set of new principles and standards that it says will guide the company’s work on artificial intelligence (AI) in the future including research and product development. Authored by Sundar Pichai, the guidelines tackle the Mountain View, California-based search giant’s objectives for AI applications and the applications that it will not pursue.
Pichai stated that Google will take the impact and benefits of AI on society into consideration as the company works to develop AI technologies, seeing to it that the positive effects far outstrips the risks. He added that Google will also aim to use AI in order to produce accurate information while respecting the “cultural, social, and legal norms in the countries” it operates in. Google will also seek to eliminate unfair biases in AI algorithms with respect to “race, ethnicity, gender, nationality, income, sexual orientation, ability, and political or religious belief.” Pichai also said that best practices in AI safety research will also be a major factor in Google’s AI initiatives as part of efforts to prevent risks by testing AI technologies in closed environments and keeping track of their post-deployment operations if need be. Google will also include privacy standards in its AI technologies and maintain transparency in how it manages data. Above all, the search giant aims to maintain human direction and control over its AI systems. Google finally promised not to develop AI for applications that it believes will lead to danger, though it added that it may pursue AI applications whose benefits exceed the risks while upholding safety standards. Pichai also said his company won’t proceed with AI development that may likely be used to spy on people or violate human rights.
The guidelines were released amid controversies surrounding Google’s AI drone project called Project Maven conducted in collaboration with the Pentagon, which is expected to turn into a $250 million business. The project involves an image analysis system used to analyze images and videos captured by drones and other means with the goal of identifying objects and tracking their movements. In April, several Google employees wrote a letter asking Pichai to end his company’s involvement in the project.