Modeling Study Shows Weak AI Rules are Likely to Backfire

New research suggests that poorly designed artificial intelligence regulations could have unintended consequences, potentially making AI products less safe than if no rules existed at all. 

With comprehensive federal oversight of AI still lacking in the U.S., state governments have introduced their measures to address safety concerns. However, experts say it remains uncertain how these varying requirements influence the decisions companies make when developing and deploying AI systems. 

To explore the issue, researchers from Carnegie Mellon University and Cornell University created a theoretical model examining how different regulatory approaches affect developers of general-purpose AI systems and businesses that build applications using the systems. 

Their findings, published in the Proceedings of the National Academy of Sciences, are intended to help policymakers craft regulations that encourage safer AI technologies while supporting innovation. 

The research team designed a model that allowed them to test minimum safety requirements for AI developers, downstream companies, or both. They then evaluated how those requirements influenced product safety and overall performance. 

Since AI regulation remains in its early stages, researchers noted that many proposals are still largely theoretical. As a result, understanding how different rules shape business incentives is an important step before broader policies are introduced. 

One of the study’s most notable findings involved regulations aimed only at downstream companies. According to the model, if the rules imposed only modest safety standards, the resulting AI products could actually become less safe than in an unregulated environment. 

The researchers defined safety broadly to include risks such as harmful chatbot responses or other forms of user harm. 

Under weaker requirements, companies creating foundational AI models might reduce spending on safety measures, including independent safety evaluations, because responsibility for the final product would rest primarily with businesses using those models. This creates a free-rider problem in which foundational AI developers shift much of the safety burden onto downstream companies. 

The study also identified a more promising approach. When both AI model developers and downstream businesses are required to meet clear safety standards, the model predicted better consumer protection alongside stronger financial outcomes. Shared responsibilities reduce uncertainty, allowing each company to rely on the other to meet established requirements rather than making assumptions about safety investments. 

Although the model simplifies the AI marketplace, the researchers plan to examine how existing regulations affect real-world AI development. They also hope to expand the framework to include multiple countries, competing AI companies, and regulators with different standards, reflecting the increasingly complex global AI ecosystem. 

For companies like D-Wave Quantum Inc. (NYSE: QBTS) that are involved in another frontier technology, quantum computing, the lack of a regulatory framework is a major concern given the way AI firms like Anthropic have faced ad hoc contradictory executive orders regarding their AI models because the country currently lacks a clear framework guiding the AI industry. 

About AINewsWire

AINewsWire (“AINW”) is a specialized communications platform with a focus on the latest advancements in artificial intelligence (“AI”), including the technologies, trends and trailblazers driving innovation forward. It is one of 75+ brands within the Dynamic Brand Portfolio @ IBN that delivers: (1) access to a vast network of wire solutions via InvestorWire to efficiently and effectively reach a myriad of target markets, demographics and diverse industries; (2) article and editorial syndication to 5,000+ outlets; (3) enhanced press release enhancement to ensure maximum impact; (4) social media distribution via IBN to millions of social media followers; and (5) a full array of tailored corporate communications solutions. With broad reach and a seasoned team of contributing journalists and writers, AINW is uniquely positioned to best serve private and public companies that want to reach a wide audience of investors, influencers, consumers, journalists, and the general public. By cutting through the overload of information in today’s market, AINW brings its clients unparalleled recognition and brand awareness.

AINW is where breaking news, insightful content and actionable information converge.

To receive SMS alerts from AINewsWire, text “AI” to 888-902-4192 (U.S. Mobile Phones Only)

For more information, please visit www.AINewsWire.com

Please see full terms of use and disclaimers on the AINewsWire website applicable to all content provided by AINW, wherever published or re-published: https://www.AINewsWire.com/Disclaimer

AINewsWire
Austin, Texas
www.AINewsWire.com
512.354.7000 Office
Editor@AINewsWire.com

AINewsWire is powered by IBN