In a move aimed at preventing the misuse of artificial intelligence (AI) technology, the United States and 17 other countries have signed a new agreement on AI safety. However, experts have expressed mixed opinions about the effectiveness and significance of the pact.
The 20-page document, signed by countries including Germany, Italy, Australia, and Singapore, acknowledges the need to develop measures to protect the public from potential abuses of AI. While the agreement is considered a step in the right direction, critics argue that it lacks specificity and enforcement mechanisms.
Phil Siegel, founder of the Center for Advanced Preparedness and Threat Response Simulation, described the agreement as “an agreement of intent than actual substance”. He emphasized the need for concrete procedures and regulations to ensure the security of AI systems.
Christopher Alexander, chief analytics officer of Pioneer Development Group, dismissed the agreement as a “feel-good measure” and called for enforceable rules and industry guidelines that hold real significance.
The Biden administration has been pushing for more regulation of AI, including signing an executive order last month to promote safer development of the technology. However, experts remain skeptical about the impact of the order and stress the need for legislative action.
Despite the skepticism, U.S. Cybersecurity and Infrastructure Security Agency Director Jen Easterly sees the international agreement as an important step towards AI safety globally. She emphasized the agreement’s focus on prioritizing security during the design phase of AI systems.
However, critics argue that the agreement falls short in addressing key concerns. Ziven Havens, Bull Moose Project Policy Director, characterized the pact as “bland” and lacking seriousness. He called for stronger legislation on AI, including generative AI and minor safety.
Samuel Mangold-Lenett, a staff editor at The Federalist, echoed the sentiment and emphasized the need for action on AI regulation within the United States. He criticized the agreement for being non-binding and lacking enforcement mechanisms, suggesting that more effort should be directed towards securing user data and intellectual property.
While the international agreement on AI safety is seen as a positive step in addressing potential risks, experts argue that more concrete measures and legislative action are needed to ensure the responsible development and use of AI technology.
