Policy and Legal

The introduction of artificial intelligence has been a boon to manufacturing, and the technology will continue to have a positive impact—as long as regulations are “right-sized,” manufacturers told Congress this week.

What’s going on: “Manufacturers are utilizing AI in myriad ways on the shop floor and throughout their operations,” the NAM told the House Subcommittee on Commerce, Manufacturing and Trade in a statement for the record at Wednesday’s hearing, where data was cited.

• “The diverse use-cases of AI in manufacturing suggest a need for a cautious regulatory approach to this groundbreaking technology: one that supports innovation and U.S. leadership in AI while providing context-specific, risk-based, right-sized rules of the road for manufacturers,” the NAM said.
• Giving testimony at the hearing, Siemens USA President and CEO and NAM Board Member Barbara Humpton discussed the many benefits of using AI in manufacturing and emphasized the need to ensure that AI regulations include “targeted” rather than “overly broad” definitions.

Industrial vs. consumer-focused AI: First, it’s important to distinguish between industrial and consumer-facing AI, Humpton told the subcommittee members.

• “Industrial AI is different from consumer AI,” she said. “Industrial AI uses controlled data from the manufacturing environment to help manufacturers create business value. Think better products, more efficient operations, a more prepared workforce. … AI will enable all companies—from startups to small and medium enterprises to industrial giants—to thrive in this new era of American manufacturing.”
• In written testimony, she added that “the core distinction of industrial AI is that it is trained on highly monitored data from sensors and machines, providing a more reliable foundation for training AI models.”

Simple, singular and targeted: Regulation of AI should be undertaken with a light touch and following a full accounting of on-the-books laws to prevent duplicative and/or contradictory rules, the NAM said.

• “[P]olicymakers should always review existing laws and regulations before enacting new ones, because most uses of AI correspond to tasks and objectives that industry has faced for a long time and that are thus highly likely to have already been addressed by existing laws and regulations,” said the NAM, which also referenced its first-of-its-kind AI report, “Working Smarter: How Manufacturers Are Using Artificial Intelligence,” released last May.
• “Similarly, policymakers must right-size any compliance burden associated with AI regulation,” the NAM continued. “The ubiquitous use of AI throughout modern manufacturing, as well as manufacturing’s dependence on innovation, underscore the need for rules that enable rather than hinder manufacturers’ development and adoption of AI systems.”

Protect without hindering: Congress “must advance industrial AI by prioritizing strong rules for digital trade, especially to include strong protections for source code and algorithms,” Humpton went on in her written testimony. “We encourage policymakers to build upon the success of previous U.S.-led efforts to protect intellectual property.”

• Legislators must also safeguard privacy and protect against baseless legal claims, the NAM said. “[I]t is … crucial that Congress take steps to maintain the privacy of personal data when utilized in AI contexts. … A federal standard should avoid a patchwork of state-level rules by fully preempting state privacy laws; it also should protect manufacturers from frivolous litigation.”

The last word: “The range and importance of uses of AI—transforming every aspect of the core of manufacturers’ operations—make it clear that AI has become integral to manufacturing,” said the NAM. “With the right federal policies, manufacturers in the U.S. will continue to devise new and exciting ways to leverage AI to lead and innovate and stay ahead of their global competitors.”