Programming note: Digital Future Daily will be on the ground at CES this week, providing coverage Tuesday through Friday. If you’re attending this year’s conference, drop DFD’s Derek Robertson a line at drobertson@politico.com. LAS VEGAS — The convention floor at CES 2024 might be packed with waves of buyers, designers and curious academics and journalists, but in a year when D.C. has become obsessed with new technology Washington regulators have their own significant presence on the ground. The breadth of regulators who’ve traveled here is a reminder of just how many emerging tech issues are facing Washington at once — AI is the big one, but also growing calls to regulate crypto, antitrust and competition, and geopolitical concerns about China and the global supply chain. The panels here are stacked with officials from the National Security Council, Treasury Department, Federal Communications Commission, Federal Trade Commission and others. For all the novelty of the technology, the overarching issue isn’t that different from the one facing every restless industry: how government can safeguard the public interest without, to borrow the common industry refrain, stifling innovation. That’s been particularly tricky amid the AI boom, with the pace of innovation quickening so much that the industry can hardly keep up with itself. “None of us know where this is going; it’s an explosive technology,” said Robert Califf, Commissioner of the Food and Drug Administration, at a panel this morning. Califf would know: he’s a former policy executive for Google’s parent company, Alphabet, and said it’s paramount for the government to get up to speed technologically so it can regulate AI properly. He had a fairly optimistic vision of AI as a technology that can self-improve — “[W]e also know that if you have a chance for an algorithm to learn, and it’s supervised correctly, it will get better and better over time,” he said — but he did see regulation, not just industry, as key to a positive outcome, concluding that “we have to be able to regulate them correctly.” Commissioner Rebecca Slaughter of the FTC discussed the commission’s philosophy on both regulating and encouraging the growth of AI, saying that using its antitrust tools to promote competition in the space is crucial for the technology to flourish. “What we want is to unlock the potential of this technology and have it grow in the market through innovation… and not be limited by a few gatekeepers who control access to some of the core inputs, whether it’s data or compute or other elements of generative AI,” Slaughter said. She also implicitly brushed aside criticisms that Chair Lina Khan’s FTC is hostile to the tech world in general: “If we were hostile to the technology, we would not care about gatekeeper control.” The ways in which AI has transformed government work and regulatory concerns can sometimes be a little less obvious than the extreme life-and-death stakes of putting powerful algorithms into biomedical devices. Yesterday I spoke with Kathi Vidal, director of the U.S. Patent and Trademark Office, who appeared in conversation with the hosting Consumer Technology Association’s senior vice president for political and industry affairs Tiffany Moore about “Protecting America’s Best Ideas.” Vidal described to me how the USPTO has tackled the problem of assigning inventorship for products created with AI assistance — how much AI involvement is too much, and the question of whether AI itself can be an inventor (which was seemingly answered in the negative last year when the Supreme Court declined to hear a case asserting AI inventorship). “There's already a construct we're working within, which is to determine what is the level of human contribution,” Vidal said, noting that a key concern is not just to correctly assign credit for an invention but to ensure tech-savvy patent trolls don’t make claim to AI-generated inventions that they have no intent to market or sell (similar to fears that AI might lead to courts being overwhelmed by AI-assisted nuisance lawsuits). She said that the USPTO will likely publish guidance in February that will address this problem, as well as further clarifying rules around AI and inventorship and the role of AI prompts in assigning patents. In the absence of any hard-and-fast rules for AI like those in the European Union, regulators are well aware that they’ll be expected to fill in the blanks when it comes to providing legal guidance for tech companies. Therefore, the delicate art of clarification seemed to be a theme among government officials all day. "We cannot by rule make illegal practices that are otherwise legal,” the FTC’s Slaughter said. “What we can do by rule is provide clarity up front around the contours of the laws we enforce… [which] on the back end also helps less honest businesses be deterred from law-breaking.”
| 
Comments
Post a Comment