An AI draft from the Japanese government calls for companies and organizations that utilize artificial intelligence to take steps to reduce over-reliance on the technology, according to Japanese media reports. The draft also calls on AI developers to refrain from using biased data in machine learning and requires them to record their interactions with the technology in order to make it available in case of problems.
The draft, which will be finalized by a group tasked with discussing the country's AI strategy, is expected to be completed by the end of the year. Japan is also working with other members of the Group of Seven (G7) to develop international guidelines for AI developers.
The draft specifies 10 ground rules for AI-related businesses, which include ensuring fairness and transparency to protect human rights and prevent personal information from being made available to third parties without an individual's permission. The rules also require information on how data from individuals or entities is accessed and how it is used by interested parties.
Under these rules, companies developing AI platforms, service providers utilizing the technology, and users will all be held accountable to some degree. Developers are being asked to ensure that data used for AI purposes is accurate and timely, and to take steps, to the extent possible, to ensure that information used without authorization cannot be accessed.
At the same time, providers utilizing AI will be required to warn users to avoid entering personal information that they do not wish to be accessed by third parties and to ensure that their services are limited to their intended use in order to prevent bad actors from abusing the technology for malicious purposes.