Strategic Management Explorer
Why Better Legal Protection Can Make Innovation More Collaborative
By Eunkwang Seo and Christopher P. Dinkel
Firms face a persistent tension when organizing innovation. Innovation depends on bringing together people with different expertise, but broader collaboration makes proprietary knowledge harder to protect. Employees with access to knowledge across many projects can more easily carry it outside the firm.
One common response is compartmentalization: separating inventors into independent groups so that no single employee has access to too much of the organization’s proprietary knowledge. This protects know-how, but the more firms isolate teams, the harder it becomes to recombine knowledge.
The Practitioner “So What”
Knowledge creation and knowledge protection are not separate processes. They are deeply intertwined. How well a firm can protect its proprietary knowledge may shape how freely it can connect people and ideas inside the organization.
Our recent study in the Strategic Management Journal, “Knowledge protection and modularity of innovation activities in firms: Evidence from US trade secrecy law,” asks: What happens to this tradeoff when the law provides stronger protection for proprietary knowledge? The question is pressing in today’s race for artificial intelligence talent, where companies increasingly turn to trade secret law when engineers change employers: Apple sued OpenAI in July 2026, alleging that former employees took confidential product information with them, and a federal judge has twice dismissed xAI’s trade secret lawsuit against OpenAI over hired-away staff. Such disputes focus on what departing employees take with them; our research examines how legal protection shapes the way firms organize the people who remain.
The Research
We examine changes in U.S. trade secret law, particularly state adoption of the Uniform Trade Secrets Act, and use patent data from 783 high-tech firms (1976–2017) to measure how modular—or compartmentalized—their inventor collaboration networks are.
In a highly modular firm, inventors cluster into tight groups that rarely collaborate across group lines. We gauge legal protection with a state-level Trade Secrecy Index based on a firm’s headquarters state, since courts most often apply the headquarters state’s law in trade secret disputes.
Because states adopted the act in different years, we can compare firms in states that strengthened protection with same-industry firms in states that had not yet done so.
What We Found
When legal protection becomes stronger, firms become less modular and more integrated. Adoption of the Uniform Trade Secrets Act reduced the modularity of firms’ inventor collaboration networks by 4.2 percent relative to the sample mean. Because modularity changes little within firms over time, that represents a substantial reorganization. The effect emerged about two years after enactment, as firms adjusted their structures. The changes also improved innovation quality, increasing technological impact by 15.9 percent relative to the sample mean. Technological impact captures how much inventions influence later technology. Novelty and originality did not change significantly; the gain in impact is consistent with the view that high-quality inventions tend to come from combining diverse knowledge across boundaries.
Why does this happen? When firms worry less about knowledge leaking out, they have less reason to compartmentalize. Consistent with this, the shift was more pronounced for firms with superior technologies or many nearby competitors—those with the most to lose from knowledge spillovers. It was weaker where patents, complementary assets, or lead time already protect knowledge effectively.
New hires play a key role. After protection strengthened, newly hired inventors—who do not typically occupy central brokerage positions within firms—became more central bridges between previously separated innovation communities. New hires bring valuable skills but being less socialized, are typically less trusted than long-tenured colleagues. Stronger legal protection appears to ease that trust problem, making firms more willing to place both experienced and early career hires in bridging roles. The positions of incumbent inventors, who already tend to be central, did not change significantly. Today’s AI talent disputes turn on the same line: in dismissing xAI’s claims, the judge found that asking a new hire about past work is not the same as encouraging theft. Clear legal boundaries of this kind may be part of what allows firms to integrate newcomers with confidence.
The evidence has limits. It covers only patented collaborations in U.S. high-tech firms; whether the pattern holds where firms patent less remains an open question. And despite extensive robustness checks, we cannot rule out every alternative explanation. Nor do the findings imply that compartmentalization is always a mistake: where legal protection remains weak, internal separation may still be a sensible defense for a firm’s most sensitive knowledge.
The mechanism matters as much as the question. Any operator who sees a defect can stop the line, so the questioning happens at the machine within minutes rather than in a review meeting weeks later. Taiichi Ohno’s canonical chain runs: the machine stopped — why? A fuse blew from overload. Why? The bearing was not lubricated enough. Why? The lubrication pump was not circulating properly. Why? The pump shaft was worn. Why? There was no strainer, and metal filings got in. Stop at the first why and you replace a fuse; reach the fifth and you install a strainer. That is the same discipline the third check below asks of an AI recommendation.
What This Means for Managers and Policymakers
These findings highlight an important lesson for managers: knowledge creation and knowledge protection are intertwined. How well a firm can protect its proprietary knowledge may shape how freely it can connect people and ideas inside the organization.
Practical Strategies for Managers
For managers, the findings point to several practical steps:
-
- Audit why teams are separated. Separate walls that serve an operational purpose from those built mainly to guard against leakage, which may be costing the firm valuable recombination.
- Treat legal and organizational protection as substitutes. Where trade secret protection is strong, internal barriers matter less. Because courts most often apply the trade secret law of a firm’s headquarters state, that state’s law should inform team design.
- Use new hires as bridges. Onboarding can connect new inventors to multiple teams rather than one group.
- Pair openness with clear protections. Confidentiality agreements, clear trade secret policies, and documented access practices preserve the legal protection that makes integration safer. In Apple’s suit, for example, OpenAI has pointed to Apple’s own practices—such as employees using personal cloud accounts for work—to contest the claims.
For policymakers debating trade secret law, employee mobility, and noncompetes, the study suggests that legal protection of knowledge shapes not only whether firms invest in innovation, but also how openly they organize it. That debate focuses largely on hiring; these findings suggest the law also shapes how firms integrate new hires. The debate has also shifted: the Federal Trade Commission’s nationwide noncompete ban never took effect and was removed from federal regulations in February 2026, leaving a patchwork of state laws—Washington, for example, will void most noncompetes beginning in mid-2027.
Protecting knowledge, in other words, does more than keep it from leaving the firm. It can also free firms to integrate the knowledge they already have.
Eunkwang Seo is an Assistant Professor of Management at the Spears School of Business, Oklahoma State University. His research examines how firms create and appropriate value in knowledge-intensive industries through the acquisition, retention, and recombination of knowledge assets embedded in innovative employees. Christopher P. Dinkel is an Assistant Professor of Legal Studies in the Department of Management at the Spears School of Business, Oklahoma State University. His research focuses on using quantitative methods to conduct empirical studies in the areas of intellectual property law, employment law, and law and ethics.
