IP Defense in Alliances

Zsolt F. — Managing Director

2026-10-07

People are socially and naturally motivated to seek opportunities for joint action. Similarly companies and organizations seek economically motivated partnerships, joint ventures or other M&A activities to form strategic alliances. These are indispensable for accessing specialized resources, capabilities, markets and know-how that companies cannot easily develop in-house. While critical for innovation, alliances must deal with distrust, especially concerns about losing intellectual property.

An alliance between the Chinese manufacturer and exporter of wind turbines, Sinovel Wind Group, and the U.S.-based company American Superconductor (NASDAQ: AMSC), tells a story on the failure to defend intellectual property. AMSC provided technology, a software that regulates the flow of electricity from wind turbines to electrical grids for the Sinovel turbines. In 2011 Sinovel conspired to obtain AMSC’s copyrighted information and trade secrets, reverse-engineered its software, and used it to produce wind turbines and retrofit existing ones with AMSC’s technology without paying the more than $800 million it owed and has promised. This predatory practice and the failure to defend against it wiped out a staggering $1 billion in shareholder equity through an 84% drop in AMSC’s stock price. On top of this, AMSC was forced to cut 700 jobs. Sinovel nearly destroyed an American company by stealing its IP.

In another example an American sports car maker Saleen Automotive entered into a joint venture in 2019 that turned out to be predatory. Saleen accused its partner of filing more than 500 patents based on $800 million worth of trade secrets and Saleen’s technologies, without crediting Saleen’s investors.

These examples demonstrate that an alliance can go disastrously off track. The formation of IP disputes is more common than it may seem, as they receive relatively little publicity. There are a number of causes and practices that lead to such disputes. If we were to sum up all the different ways, probably half of all alliances encounter some form of knowledge spillover, whether intentional or unintentional, making it a key concern. However, only a fraction of these cases end up in court.

Most of these conflicts are deliberately kept out of the media because the companies involved often depend on their partners for their ongoing operations. Demonstrating and attributing malicious intent can be particularly challenging and lengthy legal proceedings are costly. Moreover, victims are reluctant to admit they have been subjected to predatory actions, as this can cause substantial brand damage with other partners, shareholders and customers.

However, not all IP leaks involve outright theft or breach of contract. In many cases partners may employ subtle, intrusive tactics to access and use intellectual property in ways that do not explicitly violate contracts but fall into grey areas, taking advantage of trust.

Given the sheer number of such incidents it is not a question of whether this is happening in the business world, and it is not difficult to find economic reasons behind it. Malicious partners can exploit others’ IP to develop their own solutions, while avoiding the long, expensive and risky process of internal research and development. Their motivation is often to outmaneuver, catch up with or sideline partners who hold stronger bargaining positions.

Predatory behavior may also be driven by national interest policies, where the local partners are encouraged to capture foreign technology and contribute to their own countries, while eroding others’ advantages. We see this in many domains, including the current AI arms race.

Identifying these sorts of intentions a priori is very challenging, as they are often deliberately concealed. In some cases, companies knowingly enter such alliances when executives believe their IP is sufficiently protected - underestimating the hazards - in pursuit of immediate benefits such as market access or lower manufacturing costs.

Predatory partners can be highly adept at outmaneuvering IP defenses. While the loss of IP is already significant, there is another, often overlooked dimension of these cases. These partners not only extract valuable information, but can also learn how to better defend their own. Inadequate defense against knowledge acquisition may inadvertently strengthen the partner. This article will show why the principle of layered defense is important for this very reason.

Strategies for dealing with Predatory attacks

While traditional non-disclosure agreements and alliance contracts offer some legal protection against intellectual property theft, these are insufficient in dealing with predatory attacks. These legal agreements provide only reactive remedies for IP leaks, after the damage has already been done. Even when an NDA specifies monetary compensation for specific misconduct, legal remedies cannot fully reverse the competitive and economic consequences of an IP leak.

Another common countermeasure against this behavior is adopting more formal governance of the alliance, such as establishing a joint venture, or simply avoiding alliances with direct competitors who could more effectively exploit the company’s IP while restricting the scope of the cooperation to minimize the touchpoints through which sensitive information may leak.

Such preventive measures are typically determined during the contracting phase at the launch of the strategic alliance, but they offer only limited defense and flexibility once the alliance is underway.

The following sections will outline defensive techniques that can be implemented alongside legal and contractual measures. While the article does not intend to train anyone in industrial espionage and the defenses may appear general, companies such as Apple, Tesla and TSMC often use the same practices in their collaborations. Avoiding the intrusive practices can help prevent IP theft and ensure long-term competitive advantage, whether for a small startup or a large technology company.

Table 1: Intrusive and defensive techniques

1. Designating Gatekeepers

Predatory partners often nurture close relationships with key employees to foster interpersonal trust. Such ties may form during frequent informal meetings, through mutual participation in workshops and team-building activities, or casual after-work gatherings. These relationships are purposefully cultivated rather than naturally evolving and can lead unsuspecting employees to lower their guard and willingly give away sensitive information.

To counter this phenomenon, a company can restrict information channels and designate gatekeepers, employees who are trained to manage the flow of sensitive information. This role requires a combination of supervisory and diplomatic skills. Serving as a central point of contact, the gatekeeper ensures that only preapproved and necessary information is shared with the partner.

In addition to controlling other employee’s interactions with the partner, a gatekeeper regularly meets with the partner’s managers to set joint targets and exchange essential information. The gatekeepers are trained to recognise and intervene when partners attempt to cultivate manipulative relationships or exploit interpersonal trust.

For example, the gatekeeper can ensure that employees with access to critical and vulnerable knowledge, such as trade secrets and confidential plans are deliberately shielded from unnecessary contact with the partner. Meanwhile employees whose work involves secured IP, such as patents, can participate in tightly controlled collaborative activities with the partner.

Many strategic alliances involve extensive collaboration in areas such as sales and marketing, while companies often shield their engineers and engineering departments from direct involvement.

A notable example is the Renault - Nissan joint venture, which involved close collaboration in procurement, marketing and other business functions, but kept their engineering departments apart. Instead of giving the partner unrestricted access to their engineers, they implemented selective engineering exchange programs, where only carefully vetted engineers were allowed to spend brief periods at the partner’s premises. Even then, the gatekeeper monitored both of those engineers and the knowledge they might access.

The gatekeeper shall be an employee with vested interest in the company’s success, such as a long-term veteran with an equity stake or strong socio-emotional attachment to the organization. Ideally, this is someone loyal and protective of the company’s interest and less likely to disclose sensitive information. For example, some companies assign senior executives with over twenty years tenure to serve in this role, ensuring confidence in the gatekeeper’s trustworthiness and commitment to protecting the organization’s interest.

2. Cyber Defenses

Predators may attempt to bypass a company’s IP defenses through hacking, social engineering, or by exploiting vulnerabilities in its ICT infrastructure. As more alliance activities take place virtually or remotely, the importance of cybersecurity countermeasures continues to grow. However, alliances often rely on systems, third-party collaboration platforms that may have known vulnerabilities. In such cases, a company cannot rely solely on its own cybersecurity measures.

Typical defenses include encrypting of sensitive documents at rest, using tracking systems to monitor who accesses them, and ensuring confidentiality through secured communication channels and firewalls. For example, one company used secured encrypted USB drives, encrypted files, and a tracking system for both digital documents and their physical copies. Anyone who printed a document had their name automatically added to the printout, holding them accountable for its confidentiality.

Such measures are critical in alliances where multiple parties access shared resources, which significantly increases the risk of unauthorized distribution of sensitive information.

Companies shall invest in advanced security measures, such as multi-factor authentication, machine learning based anomaly detection systems, and routine security audits. Anomaly detection can identify unusual access patterns and intercept breaches before they cause damage. Cybersecurity training helps reduce the likelihood of employees falling victim to social engineering attempts, which are one of the most common tactics used by predatory partners to gain access to companies’ off-limit systems.

Companies such as Samsung, NVIDIA, and TSMC conduct internal drills where expert employees simulate cyberattacks in an effort to breach the company’s digital defenses. Meanwhile, other expert teams are tasked with monitoring, detecting and thwarting these simulated attacks. These exercises enable companies to identify and remediate vulnerabilities and loopholes before they can be exploited by real malicious actors.

3. Physical Barriers

When a partner’s employees work at or visit the company’s sites, they may engage in covert observation. This can range from inadvertently overhearing conversations to deliberately using recording devices to capture confidential information in R&D labs.

Enforcing strict physical security protocols protects critical information from both inadvertent or deliberate leaks to unauthorized individuals. Measures may include holding partner meetings in secluded areas of the building, using secure, separate offices for sensitive work, actively monitoring site access, and inspecting employees and their bags when entering or leaving the premises.

For example, Tesla and Panasonic operate Gigafactory Nevada as part of an alliance in which Panasonic supplies batteries for Tesla’s electric vehicles. The factory features strictly separated zones accessible only for authorized partner employees. All semi-finished products are transported by robots between walled-off areas to prevent human interaction across zones.

4. Process Fragmentation

Partners may attempt to recreate a product or process by gathering information across multiple touchpoints and functions. They can then piece together partially shared or incomplete information with publicly available sources, such as patent filings and technical reports.

One effective mitigation strategy here is to divide the manufacturing process into physically and geographically separate stages, or to work with different partners for different components or production phases. This approach restricts information flows and reduces the risk of any single partner gaining a comprehensive understanding of the product’s design.

This is why Apple maintains a strict practice of compartmentalization. The iPhone’s CPU is manufactured by TSMC in Taiwan, the display by Samsung in South Korea, and the final assembly is handled by Foxconn in China. When ending a partnership, Apple retains ownership of critical IP and shuts down associated facilities to prevent former partners from passing sensitive information to third parties.

Another example involves a chemicals company that entered into a joint venture with Apple to produce the glass for the iPhone’s screen. Apple disclosed only the bare minimum information necessary for manufacturing the glass within a joint venture facility. Upon termination of the partnership, Apple dismantled the entire production facility to ensure that its partner could not repurpose the glassmaking knowledge in other projects.

5. Obfuscation of IP

Strategic alliance partners often gain access to prototypes, software and manufacturing machinery that would otherwise be inaccessible to outsiders. Predatory partners can exploit this access to reverse engineer a company’s product.

To combat this, companies can adopt product and design strategies that make reverse engineering difficult or prohibitively expensive. For example, adding non-essential decoy elements, deliberate inaccuracies or design complexities to components or final products can obscure critical aspects of the product, mislead engineers attempting replication, or alert the company about reverse engineering attempts.

Some manufacturers go further by installing temporary components or circuity elements that hinder the intended functionality of the product. These elements are removed only during the final stage of production or upon delivery to the customer, without partner’s involvement.

For example, some companies such as the Dutch firm ASML, which sells advanced chip manufacturing equipment, install kill-switches that allow them to remotely disable their machinery if tampering is detected.

A company may deliberately leak outdated or less valuable knowledge to outsmart partners who may gain access to off-limits knowledge by synthesizing information from multiple available sources. By directing the partner’s interest toward less critical innovations, the company can more effectively safeguard its most valuable technologies.

One multinational tech company, for instance, shares technology that is considered outdated from its perspective but still appears new and advanced to its partners. This allows those partners to benefit from the knowledge transfer, while the company remains two or three steps ahead of them, and limiting the potential damage from leakage.

A variation to this technique is known as the “canary trap”, which uses deliberate leaks to identify protection loopholes and untrustworthy partners.

In the past, following a series of internal leaks, Apple reportedly provided several employees with slightly different versions of seemingly sensitive information. When one of those versions was leaked, the company was able to identify the source based on the unique details.

By sharing altered versions of non-critical information with different partners or employees, a company can trace how the information is handled and pinpoint the party responsible for its misuse.

6. Employee Controls

A partner may attempt to recruit and hire key employees from a company by offering more attractive compensation. If successful, they may then incentivize or socially pressure these hires to disclose their insider knowledge about their former employer’s intellectual property.

To mitigate talent poaching and the unintended transfer of IP, companies can implement strict HR policies and employee controls. These may include measures such as non-disclosure and non-compete agreements, controls on social interactions, and restricted access to systems containing sensitive information.

Monitoring compliance through regular inspections and electronic surveillance methods can deter unauthorized knowledge sharing.

Additionally, aligning long-term employee incentives with the company’s strategic goals can discourage the disclosure of critical know-how even after an employee has left. For example, when employees understand that their post-employment benefits (ie. deferred compensation, royalties or stock options) are tied to maintaining confidentiality, they are less likely to share sensitive information.

Companies also use process fragmentation across their workforces, isolating teams that work on different aspects of a project to reduce the risk of any single employee having too much critical knowledge that they could leak.

For example, at a chemicals company, as part of the manufacturing process, one team knew how much pressure needed to be applied to a particular component, another knew the temperature that needed to be maintained, and a third one knew the ratios of raw material mixtures. Only a few trusted employees were given access to the full set of data necessary to oversee the entire process.

Another example is TSMC, which enforces rigorous employee controls to safeguard its IP within its manufacturing plants. The company has a dedicated cybersecurity team that screens and trains personnel on handling sensitive information and resisting social engineering and phishing attempts. TSMC also imposes strict penalties on employees who violate non-disclosure agreements.

When defensive practices benefit the partner?

Predatory partners are continuously improving their defenses by learning practices from their counterparties. For companies, this not only results in the loss of their IP but also in unintentionally equipping competitors with long-term advantages, when the company is playing defense and it is losing the game. As a result, today’s predator is less likely to be tomorrow’s prey.

While parties frequently assume the role of either predator or defender, they often assume dual roles - simultaneously protecting their own IP while learning from their partners. In a continuous arms race, roles can alternate as their industries evolve and competitive dynamics shift.

For example, aggressive operational practices designed to prevent proprietary information leakage may have originally been observed and adopted from alliance partners. Some companies that once learned these techniques are now industry leaders, making it increasingly difficult for their former counterparts to catch up. This dual learning effect during past alliances can prevent erstwhile leaders from reclaiming their former competitive positions.

Conventional defenses aimed at restricting IP loss, such as reducing the scope of an alliance or avoiding partnerships with direct competitors often fail to prevent dual learning.

Paradoxically, when the scope of activities and touchpoints within an alliance is limited, the diversity of exchanged information is also restricted. This can actually help the partner sharpen their own defensive capabilities, but only if they manage to outsmart their counterpart. Otherwise just understanding defensive practices is insufficient for successful learning and implementation.

Interestingly, partners who already operate within the same knowledge domain can easily decipher IP are not necessarily the ones who pose the greatest threat when it comes to learning defensive techniques. Rather, it’s those from less familiar domains, so they have to invest more effort to understand the IP, who learn the most in the process of trying to overcome a company’s defenses.

Dealing with predators

This section serves as a warning for managers who are trying to understand the paradoxical nature of the dual learning principle.

Managers shall be vigilant when entering into an alliance and proceed under the assumption that the partner may be seeking access to the company’s IP. While alliances are cooperative by nature, partners often conceal a competition for ideas and knowledge. Recognizing predatory intentions and intrusive techniques is the first line of defense.

Unfortunately, by making only weak or superficial efforts to guard against such a behavior, a company may end up feeding the predator, making it stronger and better prepared to defend its own IP in future alliances.

Recommendation 1: Adoption of dynamic, layered defenses

As previously explained through the paradox of the dual learning principle, a predatory partner can only learn protective practices if they can successfully outsmart them. By implementing comprehensive, state-of-the-art defenses, a company can prevent both IP theft and the unintentional transfer of safeguarding techniques.

Managers shall dynamically combine defensive measures that fend off predators rather than feed them. Relying on too few defenses is risky, as it may allow a partner to circumvent them and learn how to replicate or apply them independently.

To prevent this, organizations shall layer multiple defenses, for example by combining obfuscation techniques with digital safeguards and by deploying gatekeepers alongside with strict employee controls.

Recommendation 2: Transformation of reactive defenses into continuously evolving practices

Companies shall consider this recommendation as an opportunity to innovate and improve their IP protection practices.

This can be achieved by establishing a specialized alliance function with a team of experts responsible for improving alliance management and IP protection practices. This team shall bring together executives with extensive alliance management experience (some of whom may also serve in the gatekeeper role), legal advisors, cybersecurity professionals and HR specialists who can develop and assign training programs on information security and align employee incentives with the company’s strategic objectives.

The team’s responsibilities shall include monitoring and implementing IP protection measures, drawing insights of previous partners’ defensive practices and continually strengthening the company’s own defenses to stay one step ahead. When done effectively, this approach not only mitigates risks but transforms potential threats into catalysts for long-term improvement.

Recommendation 3: Selective sharing of non-critical IP

When dealing with a highly skilled predator, even the most sophisticated protection practices have their limitations. Managers shall strategically share non-critical IP rather than excessively restricting the flow of information, which can hinder joint value creation.

Sharing non-critical knowledge can be a form of calculated risk, one that is less harmful than unintentionally teaching the predator how to bypass defenses and strengthen its competitive position in the future.

By carefully controlling what information is shared with the partner, the company can maintain a competitive edge through its superior understanding of its own IP and its greater capacity for innovating on it.

Finally the battle for intellectual property protection within alliances is an ongoing arms race between predators and defenders. Retaining a strong competitive position requires constant vigilance and continuous improvement of IP protection practices.