The eSafety Commissioner in Australia has given approval to an industry code that mandates search engine operators to implement safeguards against potential risks associated with generative AI.
The code, which has been revised and registered after earlier concerns, primarily aims to prevent the return of class 1 material, including child sexual abuse content generated by AI, in search results. Additionally, it mandates that AI integrated into search engines should not be used to synthesize or create synthetic versions of such harmful material.
Earlier this year, the eSafety Commissioner declined to register a previous version of the code after Microsoft and Google announced their intentions to incorporate generative AI into their respective search engines. The revised code now includes provisions that require search engine services like Google, Bing, DuckDuckGo, and Yahoo to implement measures preventing the misuse of AI to generate illicit content.
Julie Inman Grant, the eSafety Commissioner, acknowledged the rapid growth of generative AI and the need for updated safeguards. She commended the industry for delivering a code that upholds community safety standards.
The revised code is expected to take effect six months after registration. Additionally, two more industry codes are in development, focusing on designated internet services and electronic services, with the aim of enhancing protections against inappropriate content targeting children.
