A group of AI integration industry leaders today submitted a joint letter to the US House of Representatives Artificial Intelligence Task Force calling upon them to redefine roles and responsibilities for AI actors across the value chain. The letter was submitted to the House of Representatives leadership by Alteryx, Salesforce, Twilio, Box, Kyndryl, and Peraton.
The Honorable Jay Obernolte
Chair, House Task Force on Artificial Intelligence
U.S. House of Representatives
1029 Longworth House Office Building
Washington, DC 20515
The Honorable Ted Lieu
Co-Chair, House Task Force on Artificial Intelligence
U.S. House of Representatives
2454 Rayburn House Office Building
Washington, DC 20515
Dear Chair Obernolte and Co-Chair Lieu,
We represent leading companies working to integrate AI models and systems into platforms and applications that enable other businesses to innovate to the further benefit of their customers. AI is a transformative technology that has the potential to unlock enormous promise for business and consumers. We believe that effective AI policies begin with accurately identifying roles across the AI value chain and assigning responsibilities accordingly, and we believe a new approach is needed.
Current legislative discussions have centered on whether an entity is a “developer” or a “deployer” of AI. However, many of the companies leading AI innovation, including ours, often may not develop our own generative AI models, or have direct relationship with the end-user consumer. Instead, our companies commonly occupy the role of AI integrators. It is essential that any broad-based AI regulation recognizes the functional roles of different companies in the digital ecosystem before defining and assigning new legal obligations.
AI systems are developed, integrated, combined, and deployed through a complex value chain in which multiple organizations may make design or integration decisions that impact the safety or risk of the system as it is deployed in the marketplace. There are multiple actors in the AI value chain in a spectrum of roles and relationships, and their responsibilities and obligations should differ in accordance with their roles. Indeed, one company may often occupy multiple roles at once. However, AI policy debates have generally not reflected this complexity. For the US to reach its full potential on AI innovation, it is important that any AI regulation clearly and appropriately defines those roles and obligations. Yet, current policy debates lack the nuance necessary to capture this complexity and ignore key components of the value chain: namely, companies that retrain, modify, fine-tune, combine, and integrate AI models into products and services, some of which may be further modified by others further along in the value chain.
We urge the Task Force to focus on the concerns of companies heavily involved in AI who are primarily integrators, as well as developers or deployers, of AI models. Without this focus, the US could fail to capture the important contributions of integrator companies who support the needs of businesses across the US and the globe. As your work continues, we would welcome the opportunity to provide more information about the unique roles of AI integrators, our concerns about risk in the AI value chain, and specific proposals for addressing that risk including through legislative language.
Thank you for your consideration, and we look forward to working with you to address these issues and ensure that businesses can continue to lead responsible global AI innovation.
Sincerely,
Alteryx
Salesforce
Twilio
Box
Kyndryl
Peraton
cc: Members of the Task Force on Artificial Intelligence
Read press release