Ireland Degges
The Biden-Harris administration has received voluntary commitments from seven major artificial intelligence companies to collaborate to ensure that AI development is safe, secure and transparent.
Participating companies – Amazon, Anthropic, Google, Inflection, Meta, Microsoft and Open AI – will follow three major objectives to align their AI systems with standards set by the White House, the administration said on Friday.
To embrace the full potential of AI, the Biden-Harris administration has urged the private sector to make sure that their products are secure to protect the rights of Americans. The White House said that companies developing these tools “have a responsibility to ensure their products are safe,” which marks the first pillar of responsibility for committed organizations.
In doing so, these companies will put their AI technologies through internal and external security testing partially conducted by independent parties before they are publicly released. They will also share information on AI risk management with other industry organizations, government, civil society and academia.
Enterprises committed to the standards will also take on a strong cybersecurity focus based on heavy investments and insider threat protection to secure proprietary and unreleased model weights, according to the White House. These organizations will additionally direct third-party identification and reporting of any vulnerabilities found within their AI systems.
A major aspect of the commitment is to build trust within the public by implementing a variety of safeguards, such as marking AI-generated content, publicly reporting system capabilities and limitations, researching societal risks and creating advanced AI systems that can be used positively to tackle major societal challenges.
The Biden-Harris administration’s establishment of these commitments falls in line with a number of recent moves to put standards for responsible AI in place, including several meetings with private sector companies and leaders to discuss AI’s future. The administration also recently released the Blueprint for an AI Bill of Rights and National AI R&D Strategic Plan, two documents that also address responsible AI.
New developments and evolving standards in responsible AI are sweeping the government contracting landscape. For a deeper look into the implications of shifting AI guidelines, ExecutiveBiz is hosting its Trusted AI and Autonomy Forum – which will convene top AI experts to discuss current challenges and priorities in the field – on Sept. 12. To learn more and register to attend, click here.
