This executive order is just the next step in the Biden administration’s larger plan for tackling the problems – and opportunities – raised by this new era of AI.
The executive order on artificial intelligence issued by the White House today is the federal government’s most ambitious attempt to date to corral this burgeoning technology – and contains numerous items of interest for employers. Here are the 10 things about the “Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence” that all employers need to know.
1. It Doesn’t Create New Law – But Directs Federal Agencies to Take Action
The executive order covers areas well beyond employment and will help shape the government’s approach to the way AI is used in just about every aspect of our daily lives. But it doesn’t create any new law impacting the private sector. Instead, it pushes federal agencies to take action over the next year – some as soon as within the next 30 days.
It does so by creating a slew of new government offices and task forces, requiring each federal agency to appoint a Chief AI Officer and help staff a new White House AI Council. These new positions will coordinate the government’s AI activities.
Employer Tip: Sign up for FP Insights on Artificial Intelligence to stay up to speed on expected agency action.
2. Employers Put on Notice Regarding AI Job Disruption
An increasing number of workers around the country are coming down with FOBO – a fear of becoming obsolete – and the executive order addresses this existential sense of dread. It sends strong signals to the private sector that they must take employee and labor rights into account as they introduce AI into the workplace, knowing that regulatory and enforcement action is right around the corner.
Under the executive order, the Department of Labor is tasked with coming up with a plan for how federal agencies can prevent unnecessary job elimination or displacement and assist workers whose jobs end up being eliminated or disrupted by AI. Further, the Council of Economic Advisors (an executive agency) must report to the White House within 180 days how it believes AI will impact the labor market – and no doubt this report will lead to further measures.
3. Job Discrimination a Key Focus
The order also requires federal agencies to take steps to prevent employment discrimination spurred by artificial intelligence tools. Specifically, the order requires the following:
- The DOL will need to create formal guidance aimed at employers to reinforce that AI cannot be used to illegally track workers or their productivity in violation of existing civil rights laws. The order specifically notes that the technology should not be used for unwarranted surveillance of workforces.
- Meanwhile, the Department of Justice must create an AI-focused plan for federal officials whose job it is to investigate and take action against employers for alleged civil rights violations.
- Investigators at the DOL, DOJ, and other federal agencies will also receive special focused training on how to properly examine alleged AI-related civil rights violations.
- The OFCCP will be required to draft guidance on how federal contractors can avoid job discrimination when using AI for employment-related purposes.
Employer Tip: Follow the 10-step playbook we provided after a Congressman on the House AI Caucus offered a roadmap for employers at FP’s recent AI Conference.
4. Data Privacy Remains a Focus
Whenever you think of AI in the employment context, you must always think about data privacy. That’s because any AI technology needs access to an immense amount of information in order to do its job well, and that information needs to be stored, transferred, and used in a secure manner. The executive order takes on this dynamic head-on:
- It calls for federal support to develop data-preserving techniques that will allow AI models to train and learn without exploiting private data and cryptographic tools that preserve privacy.
- It pushes all federal agencies to take steps to adopt more robust privacy protections for individuals.
- The order calls for the federal government itself to lead this effort, reexamining all the ways it purchases personally identifiable information and creating new guidelines to safeguard such data.
Employer Tip: Make sure your company follows these best practices for minimizing data privacy concerns when using AI in your workplace.
5. Immigration Reform to Fuel AI Revolution
One of the more ambitious aims of the executive order is to transform the immigration system to invite foreign national workers with technical expertise to the U.S. and incentivize them to remain in the country to study and work.
This would be accomplished by updating the system for issuing visas for high-tech workers. We would see a streamlined pathway for foreign nationals to obtain green cards and other approvals to work in the U.S. The order directs officials at the State Department and the Department of Homeland Security to take quick action to accomplish these goals.
Meanwhile, the order requires the DOL to gather information from private employers about where they could use expertise from foreign national workers so that immigration officials can properly adjust their methodology for approving workers to join and remain in the American workforce.
Employer Tip: Stay up to speed on the sweeping changes proposed to the H-1B process here, and sign up for our FP Insights on immigration here.
6. Copyrighted Materials to Get Additional Protection
One of the AI nightmares we often hear from employers is the worry about employees inadvertently using copyrighted information culled from generative AI products. For example, a text-generating platform like ChatGPT could provide content that was copied word-for-word from a previously published media source, or an image-generating platform like Dall-E could create a new piece of creative work that “borrows” heavily from existing artwork. An employer reproducing these protected materials could get caught in legal hot water or suffer reputational damage
The order calls on Copyright Office leaders to step up and provide recommendations to the White House on how to protect the source material and prevent such issues from occurring. The order also asks government officials to develop guidelines on how tech companies can responsibly and legally train their AI models without violating copyright protections. It also calls on officials to create a new set of standards on “watermarking,” which is the process by which companies can label content that is created by AI and offer a layer of transparency that will help hold interested parties accountable.
Employer Tip: Adopt and enforce a policy on the use of generative AI in the workplace to prevent such dangers. You can download a complimentary sample policy on the acceptable use of generative AI tools here.
7. Manufacturing Sector to Get Boost
Recognizing that high-tech manufacturers will play a key role in developing the components that will power the AI revolution, the executive order provides a boost to the semiconductor manufacturing industry. The Commerce Department will be required to ensure that small businesses receive their fair share of the government subsidies to be paid out under the CHIPS and Science Act so that chip manufacturing is not consolidated among just a few large companies.
These smaller companies will also be in line to receive mentoring through formal government programs, access to workforce development programs, funding for physical expansion where necessary, and a direct line to vast amounts of datasets to help them remain competitive.
8. Financial Services Businesses Will Get Additional Protections
Given the sensitive nature of their business, one of the industries most concerned about potential cybercrimes and data security in the AI era is the financial services sector. The executive order requires the Treasury Department to report on ways that the financial services sector can best manage cybersecurity now that artificial intelligence presents new and broad risks. Look for that report by the end of March 2024.
9. Big Tech Gets Its Wish – Regulation is Coming
Last month, Senator Chuck Schumer asked 20 of the nation’s AI industry leaders at the first Senate AI Forum whether government needs to play a role in regulating AI, it was reported that every single person raised their hands. And now, those leaders are starting to see their wish fulfilled. Under today’s executive order, Big Tech companies will soon be required to follow a series of protocols in order to be held accountable in the use of AI.
- The Commerce Department will create a set of rules to force those companies creating largescale and powerful AI platforms to regularly report to the federal government on the steps they are taking to protect their technology from foreign enemies and other bad actors.
- If foreign actors seek to use American-based technology to help train their AI models, the companies would need to report this to the government so federal officials can keep track.
- Big Tech will need to “red team” – essentially implement critical safety tests – before rolling out new AI products. And they’ll also need to inform government officials ahead of time.
- Meanwhile, the Federal Trade Commission will be tasked with coming up with a set of rules to ensure that antitrust concerns do not raise their head in the AI space, requiring federal officials to examine and eradicate anti-competitive behavior in the marketplace. The ultimate goal of such rules would be to not only protect smaller businesses but the consumer as well.
10. What’s Next?
This executive order is just the next step in the Biden administration’s larger plan for tackling the problems – and opportunities – raised by this new era of AI. And again, it can’t create any new laws for private sector businesses on its own. It needs federal agencies to implement rules to carry out its aims. Look for them to begin dropping as soon as next month and throughout most of 2024.
The executive order calls on Congress to act as well. The order specifically asks federal lawmakers to pass bipartisan data privacy legislation – which has been a goal for several years now without any imminent progress. While we might see Congress continue its push towards AI legislation, you shouldn’t expect to see anything dramatic from Capitol Hill anytime soon.
Source: Benjamin M. Ebbink, Wendy Hughes, David S. Jones, Richard R. Meneghello, and David J. Walton from Fisher Phillips, LLC
From WCI's HR Answers Now ©2023 CCH Incorporated and its affiliates. All rights reserved.
Tags: Employers' Blog Posts