AI Regulation: A Mandate for Management

Pam Sornson, JD

March 5, 2024

On October 30, 2023, President Joe Biden issued an Executive Order compelling the development of new safety and security standards for Artificial Intelligence (AI) technologies. By doing so, the President acknowledged the role AI is already playing – and will continue to play – in the nation’s economic, industrial, and social development. He urges interested and invested parties to protect the community and promote fair and reasonable ideals when adopting its use. The Order represents just one aspect of the current global push to gain control over the ethical use of AI.

Organizing Innovation …

Regulating AI technology presents a unique set of challenges that must be addressed if the digital asset is to be safe to use and add more value than hazard to the community. A myriad of individual AI opportunities coalesce into a constellation of concerns raised by the programming. Each iteration and issue individually requires a dedicated regulatory response; together, they present a massive mission to the regulatory agencies responsible for AI’s oversight. The challenge is to coordinate the global effort to put guardrails around the tech to optimize its assets without unnecessarily impinging on its capacities.

… To Enhance the Good …

Just a short list of adopted uses for AI reveals the scope and extent of rules needed to maintain its integrity:

Connectivity – Assets included in AI programming are already connecting services and their providers to create an enhanced team capable of more together than each offers individually. In the healthcare sector, as an example, AI-powered programs are already conducting triage functions, issuing preliminary diagnoses, and sharing critical health information with far-flung medical team professionals at a pace unmatched by traditional collaborative methods.

Energy Management – The ever-expanding ‘distributed’ energy sector is also embracing AI opportunities to improve performance and build in reliability. Traditional community-wide power systems used centralized power stations to direct energy resources to their customers. Over the past few decades, however, adding in-home power options (solar, wind, and geothermal) resulted in a reversal of power flow. These resources now feed energy into the shared grid, and owners experience reduced energy costs or even receive revenue checks for their contribution to the network supply. The consequence is a power system that is vastly more complex than the original version, and the new iteration requires much more hands-on control. Emerging AI programs promise to streamline and coordinate that evolution.

Logistics – The COVID-19 era demonstrated the critical role that logistics play in the global economy, as supply chains failed, leaving millions of people and businesses without the goods and services they needed. AI in this sector is revolutionizing supply chain management and control. Automated warehouses populated by robots that don’t eat or sleep now provide a significant proportion of the physical labor involved, while sensors and cameras track the movement of goods through the system from their original creation to their ultimate destination. The efficiency level rises impressively in AI-enhanced facilities, which promises even more innovation to further the adoption of AI programming into the industry.

Each of these sectors involves thousands of companies and millions of people, all of which can influence the integrity of the respective system. Without regulation, the actions of any one person or entity can generate disastrous consequences for the others.

… And Manage the Bad.

Another short list of AI realities reveals the threats that an unregulated resource poses:

Job Losses – Automation – using robots to perform functions previously managed by humans – has already caused millions of job losses. In 2023, 37% of businesses responding to a New York-based survey stated they had replaced human workers with automated, AI-driven programming. Customer service workers were at the top of the list to be eliminated, followed by researchers and records analysts. Almost half (44%) of survey respondents indicated that more layoffs were likely in 2024.

Disinformation – The term ‘fake news‘ became ubiquitous in the past decade as political entities sought to control voters using false information to influence their actions. Using AI to produce and share inaccurate or manufactured data creates fundamental challenges for every entity that relies on its accuracy and veracity to maintain its credibility. When taken to extremes, the false commentary amounts to propaganda, which some assert is the ‘world’s ‘biggest short-term threat.’

Security – Data security – keeping personal, corporate, and governmental information safely secured behind impregnable walls – has been top-of-mind in most industries for a long time, and vast magnitudes of rules and regulations have evolved to protect it. AI presents a novel iteration – it generates ‘new’ information that may or may not be accurate or truthful. The computer program that created this ‘new’ information isn’t ‘aware’ of or concerned about its veracity (computers don’t ‘think’) and treats it like any other data bit within its reach. Consequently, the responses returned by the program are only as accurate and reliable as its affiliated databases. When those aren’t reliable, the AI response won’t be trustworthy either. Researchers note that AI software can be ‘gullible’ and produce manipulated responses when fed leading or misleading questions. It is also corruptible. Programmers can manipulate AI’s function by ‘poisoning’ its databases with false data. The action trains the technology to respond according to those directives and not pursuant to legitimate AI functioning. Without focused interventions, AI has the capacity to perpetuate biases and inequities when those influences are already programmed into its information stores.

AI presents an infinite number of iterations and permutations that can be used for good – or evil. Without any sort of regulation on its use, its misuse poses a significant threat to virtually every corner of the global community. And, with so many entities now contributing to the AI universe, both legitimate and illicit, governing bodies are appropriately focused on putting guard rails on all AI efforts in an effort to maintain global stability and industrial reliability. The President’s Executive Order is one step toward gaining control over this emerging technology that offers so much promise for a healthier, more productive future for the world.