AI regulation is on the horizon
There are indications that AI regulation is on the horizon. Fortune 500 companies are preparing themselves for the potential consequences. The current state of artificial intelligence regulation in the U.S., or the absence thereof, is a matter of great concern for Fortune 500 companies as they embark on AI projects. Approximately 27% of companies mentioned the potential risks of AI regulation in their recent filings with the Securities and Exchange Commission. This indicates a growing awareness of how AI rules may impact businesses.
The concern highlights the contrast between the European Union’s AI Act and the various AI initiatives being developed by U.S. states. Despite this regulatory landscape, corporations are still moving forward with the use of AI. The U.S. government has yet to implement regulations on AI, despite lawmakers at the recent Democratic National Convention expressing their commitment to pursuing a breakthrough in this election year.
“A combination of industry and consumer self-regulation, along with formal regulation, is necessary,” stated George Kurian, the CEO of NetApp, a data storage and services company. “When regulation is aimed at promoting the secure and effective utilization of AI, it can have a positive impact.” However, as stated in NetApp’s annual filing with the SEC, any significant delays or obstacles in the adoption of AI could potentially impact the demand for their products and may not align with their forecasts.
Other companies have varying opinions on AI regulation. According to a recent analysis by Arize AI, a startup specializing in monitoring AI models, it has been found that 137 of the Fortune 500 companies have identified AI regulation as a potential risk factor in their annual reports. These risks include increased compliance costs, potential penalties, decreased revenue, and the possibility of AI models not complying with regulations.
Motorola Solutions expressed concerns about the potential challenges and costs associated with adhering to AI and data usage laws. The company highlighted the potential for increased expenses and liability due to inconsistencies in regulations across different jurisdictions, as stated in its annual report. Motorola expressed uncertainty regarding the potential impact of new laws on its products and services. The company is concerned that increased costs or liability could potentially reduce the appeal of its AI and data products for customers.
According to Visa’s annual report, the unpredictable nature of regulatory developments and the lack of a unified global framework for AI regulation contribute to an atmosphere of uncertainty. Nevertheless, numerous corporations are not halting their AI initiatives, particularly when their competitors are advancing with the technology. While still far from being completely resolved, certain AI regulations in the U.S. are beginning to gain momentum.
Companies and AI developers view a California bill called SB 1047 as a significant indicator of how AI regulation will unfold nationwide. According to AI model makers, the bill is expected to hinder innovation and discourage the sharing of AI technology, whether it is done freely or not. There are numerous AI bills currently being considered in state legislatures, with California alone having over 30 bills on the subject.
“The sheer number of regulations being considered by the California Senate and Assembly is truly astonishing,” commented Niranjan Ramsunder, the chief technology officer and head of data services at UST, a digital transformation services firm. Whenever bureaucracy is involved, it tends to generate more compliance work and hinder productivity.
Moody’s expressed a similar perspective in its annual report, highlighting the potential consequences of utilizing generative AI. They cautioned that this could subject the credit rating company to increased regulatory scrutiny and litigation, as well as divert senior management’s attention from other important business matters.
In its annual report, healthcare giant Johnson & Johnson highlighted the global trend toward more comprehensive and nuanced regulation in the field of AI. The company noted that both the White House’s Executive Order on AI and the EU AI Act are indicative of this trend, as they call for “compliance developments or enhancements.” Although regulation can lead to higher costs and increased liability, certain companies argue that imposing restrictions on AI model makers could have positive outcomes.
Booking Holdings, a travel technology company, highlighted in its annual report the potential issues that could arise from training AI on biased, outdated, or inadequate data. The company emphasized the importance of responsible AI development and governance practices to mitigate these risks, whether they are implemented by Booking Holdings or by third-party developers or vendors.
Certain companies are proactively establishing their own AI guidelines in an effort to stay ahead of potential regulations. Bhavesh Dayalji, S&P Global’s Chief AI Officer, expressed confidence in the company’s ability to navigate industry regulations, stating that their internal AI policies and practices are well-established. He emphasized that they do not expect any significant impact on their approach due to balanced industry regulation.
In its annual report, S&P Global expressed concerns that the implementation of new regulations may have a negative impact on its ability to provide AI-based products and services, thereby affecting its competitiveness in the AI market. Financial services and insurance companies experienced in dealing with regulators may find it more manageable to navigate AI laws.
Nasdaq President Tal Cohen mentioned that the exchange operator has introduced a stock order type that utilizes AI and has received approval from the SEC. This development serves as a model for potential collaborations with regulators, addressing concerns about the management of the technology in case of any unforeseen issues. However, regulation comes with a price tag. Nasdaq mentioned in its annual report that new legislation could be expensive and require extra supervision.
According to Cohen, regulating and establishing guidelines for generative AI is a unique challenge compared to other types of AI that tend to exhibit more predictable behavior. Creating generative AI involves developing methods that enable us to stay in sync with the ever-evolving path of intelligence.
Cohen emphasized that they are committed to ensuring that nothing is released until it meets the highest standards.
