Law Firms Should Stay Abreast of Latest A.I. Regulations

The White House recently secured “voluntary commitments” from seven prominent AI companies, including Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI, to address the risks posed by artificial intelligence. These companies, despite being competitors with varying approaches to AI research and development, agreed to the commitments, which aim to enhance AI safety and accountability.

The commitments include internal and external security testing of AI systems before their release. Although this is not entirely new, the pledge encourages transparency and possibly industry-wide safety testing standards. Additionally, the companies commit to sharing information on managing AI risks with each other, governments, civil society, and academia, though concerns about information protection and the potential acceleration of AI arms races exist.

The companies also pledge to invest in cybersecurity and insider-threat safeguards to protect proprietary and unreleased AI model weights. This is seen as crucial due to past leaks and the risk of foreign entities obtaining sensitive technology.

Furthermore, the commitments call for facilitating third-party discovery and reporting of vulnerabilities in their AI systems, but the specifics remain unclear. A pledge to develop technical mechanisms for users to recognize AI-generated content through watermarking is also present, despite the challenge of accurately detecting AI outputs.

The companies promise to publicly report their AI systems’ capabilities, limitations, and areas of appropriate and inappropriate use, although the extent and frequency of reporting require further clarification. Additionally, they commit to prioritizing research on societal risks associated with AI, such as avoiding harmful bias and discrimination and protecting privacy.

Finally, the companies pledge to develop and deploy advanced AI systems to address society’s greatest challenges, focusing on areas like cancer prevention and climate change mitigation.

Overall, while these commitments lack an enforcement mechanism and often mirror existing precautions, they signal progress and willingness on the part of AI companies to engage with the government and address AI-related risks proactively.