On Oct. 30, the administration of U.S. President Joe Biden published an executive order establishing new safety and protection standards for artificial intelligence.
The document comprises 26 provisions, divided into eight groups, each aimed at ensuring AI is used to maximise benefits and minimise risks. In addition, implementation of the document should place the United States at the vanguard of defining barriers to the technology and fostering continued research.
\”The executive order … protects Americans’ privacy, promotes equality and civil rights, protects the interests of consumers and workers, encourages innovation and competition, advances American leadership worldwide, and much more,\” the site says.
The document requires developers of the most powerful AI systems to share with the government the results of safety testing and \”critical information.\” To ensure safety and reliability of the technology, standardized tools and tests are to be developed.
The document notes the rise of \”fraud and deception using artificial intelligence,\” the need to create tools to detect and remediate vulnerabilities in mission-critical software, and the development of a memo on national security. The latter is intended to outline future actions in AI and security.
\”Without safeguards, artificial intelligence can further jeopardize Americans’ privacy. AI not only makes it easier to extract, identify and use personal data, but also increases the incentives to do so, as companies use data to train AI systems,\” the order states.
According to the document, irresponsible use of the technology can lead to discrimination, bias, and other abuses.
In May, OpenAI testified before the U.S. Congress and urged the government to regulate the use and development of artificial intelligence. Earlier, researchers at the company warned of the growing threat of using language models to spread disinformation.
Earlier in June, UN representatives said that AI-generated deepfakes undermine the integrity of information and fuel hatred in society.
In turn, American scientists called for countering deepfakes, created by “hostile states.”