SixDegreesAI Uses AI to Highlight Potential Antitrust Violations by Revealing Potentially Problematic Interlocking Directorates
In October 2022, the Department of Justice announced its intention to reinvigorate enforcement of Section 8 of the Clayton Antitrust Act. Section 8 prohibits directors and officers from serving simultaneously on the boards of companies that compete with one another (subject to certain exceptions). Thus far, the DOJ Antitrust Division has identified various instances of these “interlocking directorates”, leading 13 directors to resign from their corporate boards.
For investors, regulators, lawyers, and other stakeholders, identifying potential interlocking directorate violations is an arduous process
In an effort to advance the DOJ’s antitrust mission, and inspired by a Stanford Law School paper *, we assessed what it would take for a human analyst to identify these violations - and whether our AI pipelines could be of assistance to the antitrust enforcement effort.
Humans are ill-equipped to handle these kinds of issues at scale
Based on our conversations with portfolio managers and analysts, we estimate that identifying a potential list of interlocking directorates would have taken WEEKS for an analyst to complete.
To illustrate, if an analyst were interested in X00 companies, that analyst would have needed to…
manually scrape all officers/directors from X00s of corporate filings
identify X,000s of individuals
track every individual and their corporate affiliations
determine which individuals are affiliated with more than 1 company
By contrast, machines effortlessly handle this problem at scale
SixDegreesAI’s powerful discovery tools provide a superior solution. For us, identifying interlocking directorates between X00s of companies seemed like an ideal opportunity for our AI pipelines to make a difference.
The problem of “interlocking directorates” is fundamentally about “intersections”. Our team has plenty of experience identifying intersections to answer various questions. For example:
“Who’s left Tesla to work at Apple?”
“What companies supply AMD, Nvidia, Intel?”
“Which investors are heavily invested in US tech and Chinese tech?”
In the case of “interlocking directorates”, we are asking our machines to identify “intersections” across boards of directors!
The experiment we conducted:
In-line with prior studies*, we focused on the life sciences domain. Based on data from a leading financial data provider, we constructed a list of X00 publicly-listed life science companies. From there, we filtered the list down to companies with less than $1B in revenue, and above Section 8’s ~$4.5M threshold. This left us with 300+ publicly-listed life science companies. With that list of 300+ companies in hand, our AI models got to work.
In just 2 hours, our AI models identified 200+ individuals who might be violating Section 8 of the Clayton Act. This in contrast to the WEEKS that a human analyst would have needed to complete the task.
Please note: this was only based on 300+ companies in the the life sciences domain. Our machines can complete this task for every publicly-listed company, across any domain.
Visualizing the interlocking directorates:
Given the sheer size and interconnectedness of our dataset, we generated a visualization of interlocking directorates which you can find here:
https://interlocking.sixdegrees.ai
In the visualization, red dots represent people while blue dots represent companies. As such, when a single red dot is connected to two blue dots, this represents a potential antitrust violation.
Visual patterns within the graph also provide interesting insights. For example, when a large numbers of individuals sit between two companies, this may be an implicit signal of closeness between those companies (and vice versa).
Additionally, by right-clicking on any dot and selecting “Untangle”, you will see a more “structured” view of the information. In particular, you’ll see the degrees of separation from the dot you untangled.
Our discovery technology is highly generalizable for various kinds of intersections - and beyond:
This experiment was narrowly focused on “intersections” across corporate boards. But intersections can be useful for recruiting, expert networks, lead generation, and more. For example:
Who are the most connected individuals across Big Tech?
Which US petroleum companies have the most foreign suppliers and customers?
Which executives have worked at Tesla, Waymo, Uber, and Apple?
Who has experience working in machine learning, enterprise SaaS, and in Seattle?
Critically, we are able to identify intersections at scale!
But “intersections” are just the tip of the iceberg of what our technologies are unlocking for investors.
Reach out and get involved!
For investors, regulators, lawyers, and other stakeholders who may be interested in our study of interlocking directors, please feel free to reach out: team@sixdegrees.ai
Separately, we’re seeking Biotech analysts to alpha-test a new prototype. If interested, please email us: team@sixdegrees.ai
About SixDegreesAI
Our AI-native research products give investors an unfair advantage across idea generation, deep diligence, and monitoring and alerts. Our new technologies enable analysts to answer new questions, to drive new avenues of alpha generation. We look forward to releasing additional experiments and products soon. We will also continue supporting bespoke projects that leverage our differentiated discovery capabilities.
Addendum
* many stakeholders, including our team, paid attention when researchers at Stanford Law School published, “Analysis of Over 2,200 Life Science Companies Reveals a Network of Potentially Illegal Interlocked Boards.” In their paper, the Stanford Law team shed light on hundreds of potentially illegal interlocking directorates in the life science space.