In 2019, American business is poised to embrace the commercial benefits of artificial intelligence (AI), machine learning algorithms and artificial neural networks. In a report released this past spring, professional services consulting firm Deloitte expects a major increase in global AI adoption over the next two years as the technology proves itself in core business operations.

AI-powered algorithms have the ability to decrease the cost associated with uncertainty by increasing the accuracy of business-related predictions and thus improving executive decision-making. Driven by algorithmic models continuously updated by data received, these systems can learn and improve their predictive capabilities over time.

With AI capabilities emerging across economies, international antitrust authorities are becoming more interested in the potential effect of this new technological capability on market competition. In the United States, the Federal Trade Commission’s Bureau of Competition and the Department of Justice’s Antitrust Division, the nation’s antitrust enforcement agencies, recognize the potential for AI technology to enhance the competitive market environment in the era of “Big Data.”

For example, an entrepreneurial startup with state-of-the-art AI technology has the potential to disrupt an established market before incumbents can respond.

Generally, AI has the potential to enhance competition through improved target marketing and faster competitor reactions to price changes, resulting in increased competition, lower consumer prices and improved services for customers.

Yet the potential for anti-competitive behavior by AI-related technology is what keeps antitrust enforcers up at night. Of primary concern to them is the threat of algorithm collusion among competitors. By processing increasing volumes of identical external data about a market, the concern is that competitors may reach similar AI-directed outcomes.

The AI software uses machine learning to establish prices and product production levels that automatically makes decisions through the tacit collusion of their AI systems. This process strengthens already existing tacit collusion by making this environment more stable by enhancing detection and retaliation against non-participating competitors at lower levels of market concentration, thus creating increasing antitrust problems in less concentrated industrial oligopolies.

What are the possibilities of addressing tacit algorithm collusion that potentially could arise due to the “black box” nature of machine learning and artificial neural networks? As of today, there is little antitrust policy guidance as to the proper use of AI in commercial settings. While some antitrust observers have argued that companies should outright ban certain problematic features of AI software, others have called for AI software programmers to be required to implement code that restricts the possibility of such collusion occurring.

A market for such features may induce the cybersecurity industry to respond with antitrust preventive countermeasure systems, including data perturbation, masking applications and randomization software. Contrarians, however, question whether any preventative measures are practical, given that machine learning relies so heavily on sources of external stimuli.

Fortunately, the broad jurisdiction provided by Section 5 of the Federal Trade Commission Act authorizes the FTC to investigate “unfair or deceptive acts or practices in or affecting commerce” to address potential harms that may result from tacit algorithm collusion, or algorithms increasing price discrimination, or yet other unknown anticompetitive conduct.

The U.S. Congress chose not to define the specific acts and practices that constitute unfair methods of competition in violation of Section 5, recognizing that application of the statute would need to evolve with changing markets, technology and business practices. Thus, Section 5 offers no insight into what types of anti-competitive conduct will initiate agency scrutiny into this cutting-edge technology and its industry applications.

There are antitrust scholars who argue that AI is a transformative technology and will require new authority and institutional responses to effectively monitor and enforce antitrust statutes. However, before moving forward on the legislative front, it would be astute to consider existing federal antitrust enforcement agency options to meet this challenge. One such option would be establishing formal antitrust guidelines or policy statements developed and issued jointly by the FTC and DOJ.

The guidelines or policy statements, published in the Federal Register (which also solicits public comment from a variety of interested stakeholders) would provide valuable information to businesses with respect to actionable practices by the antitrust authorities concerning AI and machine learning algorithms. Moving this guidelines or policy statement process forward will take time, which allows for evidence-based antitrust policy and business practices to co-evolve.

This is certainly not the first “brave new (antitrust) world” of possible anticompetitive conduct that the FTC and DOJ have addressed. Whether it has been joint guidance on the licensing of intellectual property, or an enforcement policy statement on the health care industry, these antitrust agencies have risen to the challenge. Again, these agencies have the opportunity to provide clear antitrust guidance to U.S.-based companies as AI technology evolves and industries integrate it in their business operations in upcoming years.

Thomas A. Hemphill is the David M. French Distinguished Professor of Strategy, Innovation and Public Policy in the School of Management, University of Michigan-Flint. He wrote this for InsideSources.com. The opinions are the writer's.