Some of the world’s leading artificial intelligence companies, including Google, Microsoft, OpenAI, and Anthropic, have come together to form a new industry body called the Frontier Model Forum. The forum’s goal is to collaborate with policymakers and researchers to regulate the advancement of cutting-edge AI technology.
The companies involved stated that the forum will focus on developing best practices for AI safety, promoting research into AI risks, and sharing information with governments and civil society. This announcement reflects the industry’s efforts to establish voluntary guidelines for AI development ahead of anticipated legislation by US and European Union lawmakers.
This news follows a prior commitment made by the four AI firms, along with Amazon and Meta, to subject their AI systems to third-party testing and clearly label AI-generated content. The industry-led forum also plans to make technical evaluations and benchmarks publicly accessible through a library.
Microsoft President Brad Smith emphasized the responsibility of AI technology creators to ensure its safety, security, and preservation of human control. The forum’s initiative aims to bring the tech sector together in advancing AI responsibly to benefit humanity.
In a recent hearing, AI experts, including Anthropic CEO Dario Amodei and AI pioneer Yoshua Bengio, voiced concerns to lawmakers about the potential societal risks associated with unregulated AI development. Amodei highlighted the misuse of AI systems in areas such as cybersecurity, nuclear technology, chemistry, and biology, expressing the fear that AI could be used to create functional biological weapons.
To mitigate potential harms, Bengio recommended restricting access to AI systems, implementing standardized testing regimes, limiting the comprehension capabilities of individual AI systems, and constraining their impact on the real world.
The European Union is pursuing legislation that could be finalized this year, with intentions to ban predictive policing and limit AI use in lower-risk scenarios. In the United States, there is a greater need for comprehensive AI legislation. Senate Majority Leader Chuck Schumer plans to educate members on the basics of the industry through a series of briefings, followed by additional panels to examine the implications of AI on jobs, national security, and intellectual property.
The establishment of the Frontier Model Forum underscores the industry’s commitment to responsible AI development and lays the groundwork for collaboration and regulation in the future.
