We Get AI for Work: Unpacking the Federal Deregulation of AI

Transcript

INTRO

Biden-era guidance on responsible AI use has recently been removed from public-facing websites of certain federal agencies including the EEOC, OFCCP and DOL. However, that does not mean companies are absolved from legal responsibilities when utilizing AI in support of business operations.  

On this episode of We get AI for work, we discuss the developing trend of deregulation of AI at the federal level and what laws employers can expect to be enforced on the state and local levels.  

Today's hosts are Eric Felsberg, principal in Jackson Lewis's Long Island office, and Joe Lazzarotti, principal in the firm's Tampa office and co-leaders of the firm's AI Group.  

Eric and Joe, the question on everyone’s mind today is: How can organizations adopt and innovate AI technology and remain compliant with relevant laws and regulations, and how does that impact my business? 

CONTENT

Eric J. Felsberg
Principal, Long Island

Welcome to our latest episode of We get AI at work. My name is Eric Felsberg, and I'm here, as always, with my colleague and friend, Joe Lazzarotti. 

Joe, we have what we think is a very interesting episode today as we discuss AI and government enforcement. I think we can all agree that a lot has happened over the last month or so as our nation has transitioned to a new presidential administration under President Donald Trump. The new administration's policies have certainly impacted many parts of our lives, especially our favorite topic, which is artificial intelligence.

Joseph J. Lazzarotti
Principal, Tampa

Eric, it's good to be with you, as always. I think that's right. To say that there's been a distinct change in direction is probably an understatement. Since the inauguration, there have been quite a few executive orders, etc. AI is no different. 

There has been quite a bit of thinking around AI. How do we regulate it? Do we regulate it? I remember talking at some point about a meeting in New York that Senator Schumer was leading with some others, and there was a big question---should we even regulate it at all? Ultimately, the Biden administration began to build a framework. A blueprint, I think, was one of the documents that were put out by the White House at that time, trying to think about ways to put in place some guardrails around development. 

Then, of course, the Trump administration came in and is taking a bit of a different approach, hoping to provide an atmosphere and environment for greater innovation. In the process, beginning to pull back, but we're not really sure where the administration is headed, but it certainly looks like it's looking for more innovation in AI.

Felsberg

One of the things that we've seen from the Trump administration compared to the prior Biden administration is that Biden’s administration, as you mentioned, was very interested in establishing guardrails and putting a lot of attention on where AI could go wrong. We spent a lot of the last year or so immersing ourselves in some of the guidance we saw from that administration and how to react to it. 

The Trump administration has really moved on to this; they use the phrase ‘global AI dominance,’ which means let's make the United States front and center on this issue of AI. For starters, President Trump has overseen the rescission or the backing off of the executive orders that came out under the Biden administration and have pulled back some of the guidance that we came to know on some of these websites. 

For example, there was Executive Order 14110 Safe, Secure, and Trustworthy Development of the Use of Artificial Intelligence. That was rescinded on inauguration day. So, this was an EO that former President Biden put into place, and as you can tell from the name, the emphasis was let's move forward with this AI stuff but let's be careful and make sure that we're doing it in a safe, secure way and that we can trust the development of some of these technologies. 

Trump did away with that almost immediately upon entering office. If you haven't, to our listeners, taken a tour recently of the EEOC's website, OFCCP and other aspects of the Department of Labor website, you'll be in for somewhat of a surprise because a lot of the content insofar as AI was concerned has been removed. There was a technical assistance document that was formally on the EEOC's website that was aimed at assessing the adverse impact of software algorithms and AI in connection with their use in employee selection procedures under Title VII of the Civil Rights Act of 1964. You cannot find that anymore; it seems to have been removed from the EEOC's website. Additionally, there was a fact sheet dedicated to the use of wearables. There was this fact sheet that talked about the issue of wearables under the umbrella of federal employment discrimination laws, but that, too, has been removed. There was another document focusing on the Americans with Disabilities Act and the interplay between that act and artificial intelligence. You cannot find that. If that's not enough, the EEOC had an Artificial Intelligence and Algorithmic Fairness Initiative that is off the website as well. Not only has there been a marked change in focus by this new administration, but they seem to be just erasing, if you will, a lot of these Guidance documents from some of the public-facing websites.

Lazzarotti

I know we wrote on the EEOC Guidance on Title VII, and then there was also one on the ADA. I know you've been dealing with the OFCCP issues regarding artificial intelligence and equal employment opportunities for federal contractors. What's interesting, just as an example, is that looking at the EEOC documents on ADA and Title VII, the agency, even at that time, said this is not a binding regulation. It is just guidance and a statement about how the law might apply and some best practices. What I came away with from that document was that it's beginning to indicate what this Agency's position was and some things that you can do in advance. I don't know if those are necessarily things that you should still ignore. While the agency may be going in a different direction, at least during this administration, there is still a body of law out there, certainly at the federal level and at the state level, that companies are still going to have to grapple with. That may not be going away. 

Time is going to tell in terms of what we see in terms of regulation at the federal level anyway, but certainly, the Trump administration is going full steam ahead with looking to tout innovation. An announcement came out recently about an investment, like half a trillion dollars, which was announced with Apple after meeting with the company. That seems to very clearly align with the President's objectives with AI. The Vice President spoke at a summit in Paris along the same lines of looking to make the U.S. the gold standard, especially after the platform that China announced. 

There is all this going on with AI, and who's going to have dominance in the world of a technology that is really ubiquitous and has gotten so influential? You can see both sides of it, I suppose. The side of wanting to not have regulation constrain innovation and development and to be the leader in this technology that's so influential. At the same time, trying to say, but how do we also protect civil rights? How do we also protect privacy and data security? How do we also help people make decisions in a way that isn't going to hurt people in a bad way? So again, we'll have to see how that battle continues over the next months and years.

Felsberg

It's interesting because, to the untrained eye, people may be left with the impression that the Trump administration seems to be pulling back on the documents that we talked about earlier. Maybe the perception is, hey, then there's really very little risk here for us in terms of an AI use perspective. But if you think about what the Biden administration was doing, I wouldn't say they were necessarily coming up with new laws aimed at AI. They were just saying, look, there's this new technology that is exciting. Do not forget that there are laws that apply here, and we're going to offer you some best practices and guidance on how you can stay clear of some of those laws and still use AI. Even though all these Guidances are coming down, AI could still be used in a manner that is violative of things like Title VII, the ADA or the ADEA---Age Discrimination and Employment Act. A viable claim that is premised on discrimination that's, again, violative of those laws may still be premised on how an AI platform is being used.

One thing we've seen through a lot of the Trump Executive Orders is that they may not be giving a lot of guidance and so forth, but what's clear from the Trump administration is you better not violate federal law. If you violate federal law, that's where this administration will pay attention. If AI is being used in a way that disparately treats people, it could certainly be a violation, and you would expect some sort of enforcement.

Joe, you and I talked about how there have been some rumblings out there about the whole notion of disparate impact. This idea that you didn't mean your intent to discriminate. You had a facial-neutral policy, but the effect of that particular policy is excluding people from the employment process, for example, at disproportionately greater rates, let's say women versus men, than their counter demographic. There have been some rumblings out there about whether that theory of discrimination will go away at some point, but for the time being, it's still here. Some of these AI tools and the way that they're used could yield a disparate impact that may be a violation of laws like Title VII and others. 

Truth be told, we didn't really see a lot of enforcement or that much concern on the federal level. A lot of concern has been raised over the last couple of years at the state and local levels. I sit outside New York City, where we have the New York City Automated Employment Decision Tool law, which is really focused on the issue of bias in the use of AI tools in the city. We have Colorado's law aimed at high-risk AI systems. We have others in Illinois and California that also present compliance obstacles for those who are involved in the AI space. Some of these laws are aimed at the developers as well as the end users of the technologies.

I say all this to say that, especially to our listeners, don't be lulled into a false sense of security where you say, look, the enforcement around this is going to be lax. You are subject to these existing state and local laws. We absolutely expect more to come. Just because there's not a lot of guidance on these websites, federal law can still be violated vis-a-vis the use of an AI tool. 

Joe, with all that in mind, what would you say an employer should be thinking about now in the current environment?

Lazzarotti

Just to key off what you were saying to take a page at a privacy law in the U.S., I think for the last at least 20 or so years, the United States hasn't adopted a national federal privacy law as you see in other countries. Yet, the states moved in, and we see 19 or 20 states with comprehensive privacy laws. We've been trying to pass a federal data breach notification requirement that is generally applicable, but that hasn't happened. Every state in the country now has a data breach notification law. So, when you think about this, I certainly agree there's a level of federal law that certainly presents risks for companies, but there's also a growing body of state law. 

As you mentioned, in New York and Illinois, there's a significant set of regulations coming out likely soon in California. Both with respect to the California Civil Rights Council, which just proposed its second round of modifications on rules that get at these same issues we're talking about in terms of civil rights, discrimination and so on. One issue out of that second round is increases in record keeping that employers who use these types of decision tools have to really think about. So, there's really not necessarily a loss of regulation and law on how these technologies will be applied. The other thing to remember is that we're talking about just another layer of regulation: AI law, if you will, is going to be applied. The reality is there's a whole bunch of existing laws, whether it's treating people, discriminating against people, how we are collecting or handling their data and how we go about making decisions. Those things still have to happen. AI is just a tool that facilitates that in a way that may require additional regulation. 

To answer your question, certainly, employers will have to continue to watch these developments carefully in their state, how they're using them and where they're using them, so they understand the jurisdictional issues. That's going to be important. We see that with some of the privacy regulations with any technology, whether they’re AI-based or not, I see a lot of organizations struggling with how to understand what this tool is actually doing. What risks does it present, not just from a legal perspective but from a practical perspective, an administrative perspective and a business perspective? Do your clients want you to be using that technology? Governance is still important, and policies around collecting and using significant amounts of data are still important, regardless of where the Trump administration heads. Companies still have to think about those things.

Felsberg

Now, more than ever, you have to monitor how AI is being used in the workplace. You talked about governance, and a big part of that is, for our listeners, that if you don't already, you really should think very strongly about having a policy in place. Don't just put a one-size-fits-all policy in place but think about how it's being used in your workplace. What are you trying to control for?

That is going to be really critical in the current environment because, again, it's only going to become more onerous as we move forward. Certainly, a lot to think about. Joe, as always, great discussion. For our listeners, we hope you found the discussion helpful. As always, if you have any questions or would like us to cover a specific AI-related topic, please don't hesitate to reach out to us at ai@jacksonlewis.com.

Thanks again for listening, and we'll be back soon.

OUTRO 

Thank you for joining us on We get work®. Please tune into our next program where we will continue to tell you not only what’s legal, but what is effective. We get work® is available to stream and subscribe to on Apple Podcasts, Libsyn, SoundCloud, Spotify and YouTube. For more information on today’s topic, our presenters and other Jackson Lewis resources, visit jacksonlewis.com.

As a reminder, this material is provided for informational purposes only. It is not intended to constitute legal advice, nor does it create a client-lawyer relationship between Jackson Lewis and any recipient.