The Organization for Economic Cooperation and Development plans to review its guidelines on artificial intelligence in the wake of the precipitous uptake of generative AI, such as ChatGPT, sources close to the matter said Monday.
The OECD's legally non-binding AI Principles, first adopted in May 2019, promote the technology's use while calling for respecting human rights and democratic values.
They are intended to provide guidance to member states when formulating policy, and changes to the principles are expected to influence how member countries approach AI in the future.
The move comes as the Group of Seven leading industrialized nations has also begun motions toward drawing up concrete rules on AI, with their working-level meetings on the plans starting online Tuesday.
During the G-7 Hiroshima summit held earlier in May, the leaders agreed to formulate the group's policy on generative AI by the end of the year. The OECD's updated guidance is expected to conform to the G-7 discussions.
Since its launch in November, ChatGPT, developed by U.S. tech firm OpenAI, has captured the world's attention with its ability to quickly produce human-like text.
Along with hopes for the potential efficiencies and new opportunities the technology presents, concerns remain over privacy violations from the enormous amounts of online data the systems draw on to produce their content.
The guidelines, believed to be the first multi-national shared standards for the technology, consist of five main principles, including commitments to transparency and safety.
Among the principles that stakeholders are being asked to implement is one calling for AI-system design that respects democratic values and includes safeguards for a fair society.
From the outset, reviews of the principles after a certain period of time had been planned, a Japanese government source said, adding the revised principles are expected to reflect issues that have emerged amid the swift spread of generative AI software, capable of producing writing and generating images.
While the nature of the amendments for review has yet to be decided, the OECD will likely consider altering the language to reflect changes in the AI landscape and introducing new principles.