## OpenAI and Anthropic Sign Agreements with US Government for AI Model Testing
OpenAI and rival company Anthropic have recently signed agreements with the US government to have their new AI models tested before public release. The National Institute of Standards and Technology (NIST) will oversee the “AI safety research, testing, and evaluation” conducted by both companies. This move marks a significant milestone for OpenAI, as it is the first time the company has opened up access to third-party scrutiny and accountability.
## Addressing Safety Risks Associated with Generative AI
Generative AI has long been associated with safety risks. Its tendency to produce inaccuracies, misinformation, and enable harmful or illegal behavior has raised concerns within the industry. Additionally, generative AI models have been criticized for entrenching discrimination and biases. OpenAI has recognized these risks and has taken steps to address them internally. However, the company has been secretive about the specifics of its models and training processes.
## OpenAI’s Commitment to Responsible AI Stewardship
OpenAI’s decision to collaborate with the US government reflects the company’s commitment to responsible AI stewardship. The CEO of OpenAI, Sam Altman, has been vocal about the need for AI regulation and standardization. By working with the government and allowing third-party scrutiny, OpenAI aims to ensure that its AI models adhere to high safety standards.
## Critics’ Concerns and OpenAI’s Response
Critics of OpenAI argue that the company’s willingness to collaborate with the government is a strategic move to shape favorable regulations and eliminate competition. However, OpenAI’s CEO, Sam Altman, has emphasized the importance of national-level involvement in AI regulation, stating that the US needs to continue leading in this field.
## Building on the Biden Administration’s AI Executive Order
The collaboration between OpenAI, Anthropic, and NIST aligns with the AI executive order issued by the Biden administration in October 2021. This order mandates that AI companies grant access to NIST for red-teaming before releasing their AI models to the public. It aims to ensure the safe and responsible deployment of AI technologies. The recent collaboration announcement also highlights the intention to share findings and feedback in partnership with the UK AI Safety Institute.
In conclusion, OpenAI’s collaboration with the US government and NIST marks an important step towards responsible AI development. By subjecting their AI models to third-party scrutiny and testing, OpenAI aims to address safety risks associated with generative AI and contribute to the establishment of industry-wide standards. This partnership reflects OpenAI’s commitment to responsible AI stewardship and the need for national-level involvement in AI regulation.
