President Trump supports the growth of AI. Who protects consumers?
Sarah Meyers West of AI Now says the Trump administration is increasing federal support for AI companies while weakening state safeguards.
Monday, July 27, 2026, episode of the podcast The Excerpt: AI is already impacting decisions about health care, jobs, prices, and more, but the U.S. lacks a single federal rulebook governing this technology. Instead, oversight relies on a patchwork of state laws, voluntary standards, existing consumer protections, and post-harm litigation. Sarah Myers West, co-executive director of the AI Now Institute, joins host Dana Taylor to discuss why regulators are struggling to align, how the European Union’s approach differs, and what an effective U.S. framework would look like.
Press play in the player below to listen to the podcast and follow the transcript below. This transcript was automatically generated and edited in its current format for clarity. There may be some differences between audio and text.
Podcast: For true crime stories, in-depth interviews, and more USA TODAY podcasts, click here
Dana Taylor:
AI is advancing rapidly. Congress is not like that. In Europe, lawmakers have already passed comprehensive AI legislation, but in the United States there is no single federal rulebook. Instead, AI governance is being shaped through a patchwork of state laws, federal guidance, voluntary corporate commitments, limited FTC enforcement, and litigation when problems arise. Maybe that’s the American way. Drive innovation first, then hold courts and regulators accountable. But as AI tools move from answering questions to taking action, writing code, advising users, influencing decisions, and interacting with people in moments of vulnerability, is that enough?
Hello. Welcome to this excerpt from USA TODAY. I’m Dana Taylor. Today is Monday, July 27, 2026. Joining me to explain it all is Sarah Myers West, co-executive director of the AI Now Institute, a nonprofit, nonpartisan organization that conducts diagnostic and policy research on AI. Sarah, thank you so much for being here.
Sarah Myers West:
Thank you so much for having me.
Dana Taylor:
Before we get into regulation, I’d like to briefly touch on why many of us, including technology companies themselves, feel that regulation is so necessary. Can you explain the risks we all face because of generative AI?
Sarah Myers West:
of course. So AI systems have been in place for decades. Various forms of AI are being used in a variety of ways, ranging from highly impactful areas such as being used to decide whether health insurance companies will approve our claims, whether we will be invited to a job interview, how much the products we buy off the shelf will cost, and more. AI systems are already being deployed in many ways to shape people’s lives on a daily basis. And, just as you said at the beginning, regulators generally deal with the impact of these systems on people’s lives almost reactively. That’s when AI is used incorrectly, when it discriminates against someone, when it makes an error that could cause some kind of harm, or when it can be used for fraud or fraud. That’s the point where we improve the system. The current situation is one in which people are affected by impacts that can affect their resources and life chances, yet the deeper scrutiny required as a first line of protection has not been undertaken.
Dana Taylor:
Institutionally speaking, why can’t the U.S. government regulate generative AI like other technologies?
Sarah Myers West:
Indeed, there are already many laws in place that could be enforced against AI companies. I think a lot of policymakers are saying that there’s actually no exception for AI in the law. And I think that’s a really good starting point. But in general, I don’t think there has ever been such head-on scrutiny of this field in the way that it needs to, given the impact this field has on so many people.
Dana Taylor:
As mentioned earlier, Europe has taken a very different path than the US, with the AI Act creating an extensive risk-based framework for AI. Broadly speaking, how is Europe approaching regulation and is it working well? If so, could the US implement a similar strategy here?
Sarah Myers West:
So in the European context, regulators passed this very large regulatory package called the AI Act. And what the AI Act does, by and large, is categorize the different stages of AI use by risk. It turns out that certain uses of AI are so low-risk that they don’t really need much scrutiny. Some domains are riskier and require more attention. There are also some areas where AI cannot be used at all because it poses unacceptable risks. I think there are advantages and disadvantages to this approach. I think one of the benefits is that you can adjust it to your actual usage. One drawback, however, is that many AIs are general-purpose technologies rather than developed for specific use cases, and therefore require further scrutiny before being widely used. Although such a broad approach to the AI sector has rarely been taken in the US context, this is one path forward that some policymakers are considering.
Dana Taylor:
So, instead of the regulations we see on AI in Europe, three states, California, New York, and most recently Illinois, have enacted similar laws focused on frontier AI safety and transparency. And Colorado and Texas have very different laws. How are these states trying to reduce risk, and is that enough?
Sarah Myers West:
As a result, various states have passed laws that attempt to focus on this complex problem area. The continuum of approaches focuses on mitigating the most remote and catastrophic risks. This is the type of law passed in places like California and New York. I think one of the limitations of that approach is that by and large, most of the uses of AI that are already in place are left out of oversight. Applies only to AI systems above a certain size. It means that most of the harm that people may be feeling here and now is not getting much attention. This reduces the more catastrophic risks we face in the future. But as I said earlier, even if AI is used to deny health insurance claims, it may persist as well. And that’s an area that needs more attention.
Dana Taylor:
NIST (National Institute of Standards and Technology) created the AI Risk Management Framework, which is essentially voluntary guidance for managing AI risks. In a country without binding AI laws, is NIST still the most important venue for shaping AI safety standards? And what about the Federal Trade Commission and the Consumer Financial Protection Bureau? What role should they play?
Sarah Myers West:
As you mentioned, the National Institute of Standards and Technology has historically played an important role in technology in providing evaluation frameworks for evaluating AI systems. One such framework is the risk management framework. They are actually designed in a way that is only intended to provide information to businesses, but is not binding. There are no frictions that force companies to behave a certain way or change their incentive structures. And I think what we’re seeing now with the dynamics around AI is that a lot of companies are being allowed to grade homework on their own. And it will not meaningfully protect the front-line impacts felt by communities across the country.
Dana Taylor:
President Donald Trump took office promising global dominance through AI as well as deregulation. The administration is also putting pressure on state AI laws. Meanwhile, his chief AI policy adviser, Sriram Krishnan, resigned at the end of June. How should we understand the direction of President Trump’s AI policy?
Sarah Myers West:
If anything, the Trump administration’s stance on AI is the most extreme version of what is already a step-down approach to the field. Both seek to create deeper regulatory carve-outs that would inhibit states’ ability to use their jurisdiction to protect their voters. And similarly, the government has taken this approach of doubling down on support for specific companies, including attempting to broker international deals on behalf of certain AI companies under the AI Export Executive Order, allocating federal land to build data centers, and providing billions of dollars in loans to some AI companies to build data centers. So I would say this is an administration that has really given a thumbs up to support certain interests within the AI industry.
Can the US regulate AI without slowing innovation?
There is no single AI rulebook in the United States. This excerpt examines the risks, regulatory gaps, and discussions about what happens next.
Dana Taylor:
Does Congress need to step in to give federal agencies like the FTC and NIST the power to regulate generated AI? Is that the solution? If so, is there the political will to do it?
Sarah Myers West:
Congress certainly has an important role to play here. One is to ensure that we have the appropriate technical staff and resources to enforce the laws that are already in place, before we begin to strengthen the legal framework to enable meaningful oversight and accountability. I think one of the most significant gaps is that given the major economic and political capital that AI companies have amassed, these companies are some of the largest and most resource-rich companies we’ve ever seen in practice. That makes it very difficult for them to apply the law in a way that causes real friction. For many of these companies, FTC fines are likely to be a budget item. So I think it’s critically important that Congressional leaders not just act, but act in a way that meaningfully changes the actions of companies that have shown themselves not to have the needs and interests of the broader public in mind.
Dana Taylor:
Sarah, if AI causes harm, how do U.S. courts and regulators decide who is responsible: the company that built the model, the company that deployed it, or the user?
Sarah Myers West:
It’s still a somewhat murky area. And with the current generation of AI systems in place, it can be difficult because many of these systems are designed in a way that makes it difficult to attribute responsibility. It’s hard to definitively say, “This input led to this output. We know that it produced this discriminatory result because it was trained on this data.” Some of this is by design. These companies were publishing research that provided broader information about what kinds of systems they were building, what data they were being trained on, and what really mattered to the broader public good. Over time, they become increasingly opaque to the public and regulators, making it more difficult for courts to enforce the laws.
Dana Taylor:
And finally, what does a viable U.S. AI governance framework look like that protects people without hindering innovation?
Sarah Myers West:
I think we need to reject this fiction that regulation hurts innovation. I think under the current paradigm there is a lot of regulatory oversight that helps foster innovation. It will help fight the monopoly that has developed within the industry. It will protect the people. It also protects businesses from untested systems that may have security flaws that may not work as intended. There are many ways in which increased regulatory oversight could actually be beneficial and lead to the development of gold-standard AI by weeding out some of the bad apples developed along the way.
Dana Taylor:
Good to hear your insights. Thanks so much for joining us, Sarah.
Sarah Myers West:
absolutely. Thank you for having me.
Dana Taylor:
Thank you for your attention. I’m Dana Taylor. I caught up within 20 minutes. Come back tomorrow for another take on an important story.
