New VPA Report Recommends “Net Neutrality for AI”Vanderbilt Law School

Startups use foundational AI models from Anthropic, OpenAI, and Google to innovate in AI applications, which leads to conflicts of interest as these companies also compete with the startups.
A report from Vanderbilt Policy Accelerator’s director of AI and technology policy, Asad Ramzanali and Akhil Rajan, explores conflicts caused by these model providers competing with the startups they enable.
The market for foundation model APIs is largely controlled by Anthropic, OpenAI, and Google, who together command nearly 90% of the market by revenue but also offer their own competing AI applications.
The report suggests adopting “AI neutrality” rules akin to net neutrality principles to prevent model providers from favoring their own products over competitors’, thus protecting innovation.
The case of the AI coding agent Windsurf, whose access to Anthropic’s model was cut off amid rumors of an acquisition by OpenAI, illustrates the problematic practices these companies might engage in.
Ramzanali and Rajan propose legislation to ensure AI model providers maintain neutrality, ensuring fair access, latency, cost, and quality of service for all customers.
The Vanderbilt Policy Accelerator works on issues in political economy and regulation to bring timely research and policy proposals to maturity.

You may have missed