Amazon entra nell’US Artificial Intelligence Safety Institute per un’IA etica

In recent years,artificial intelligence has emerged as one of the most transformative technologies of our generation. The advent of large language models(LLMs) with hundreds of billions of parameters has opened up new scenarios for the use of generative AI to improve customer experience, increase employee productivity, stimulate economic growth, and more.

Amazon is committed to using this technology responsibly, working with its customers and governments to develop and deploy AI systems with safety, fairness and security at the forefront. The company supports governments’ efforts to put in place effective risk-based regulatory frameworks while ensuring that they can continue to innovate and apply the technology in practice.

The importance of U.S. regulatory leadership for AI in 2024

To ensure a comprehensive view onAI in the United States, it is critical to promote safe and responsible innovation, foster competition, and encourage policies that can be embraced by our trading partners. Against this backdrop, Amazon is pleased to announce its membership in the U.S. Artificial Intelligence Safety Institute Consortium, established by the National Institute of Standards and Technology (NIST), to promote collaboration between government and industry to advancesafe and reliable AI.

Amazon will work with NIST as part of the U.S. Artificial Intelligence Safety Institute Consortium to establish a new measurement science to identify proven, scalable and interoperable measurements and methodologies to promote the development ofreliable AI and its responsible use. It is important to note that NIST does not evaluate commercial products within this consortium and does not endorse any products or services used. To learn more about the consortium, you can visit the official website.

In addition, Amazon will contribute $5 million in computing credits to the Institute to enable the development of tools and methodologies that organizations can use to assess the security of their model bases. Specifically, the goal is to develop assessment methodologies for models with a very large number of parameters and to improve pre-deployment testing by focusing on industry-specific risks. This work, together with important ISO standards and contributions from other AI institutes, will help create a reliable and interoperable foundation for the development and deployment of responsible AI.

Institute announced by Vice President Kamala Harris

TheInstitute was announced by Vice President Kamala Harris during the Global Summit on AI Safety held in the United Kingdom. This important event brought together political and business leaders from around the world and built on the foundation of the Executive Order signed by President Biden on October 30. Earlier this year, Amazon also endorsed the White House Voluntary AI Commitments, which are based on flexible and enduring best practices that can evolve along with AI developments, setting ambitious and concrete goals to manage potential risks.

We recognize that the unique challenges posed by generativeAI will require continued cooperation, and we look forward to working with NIST and other consortium members to improve the security and stability of generative AI. Our involvement in this institute is just one of many steps we are taking to invest in the future of responsible AI and help set international standards in the interests of our customers and the communities in which we live and work.