Building deep-tech advantage in CEE: From IP strategy to global expansion (Part II)

based on interview with Dr. Dipanjan "DJ" Nag, Venture Partner at Radix Ventures, the President of Innovaito, LLC, a Professor of Practice at Rutgers University, and a globally recognized expert in intellectual property strategy

‍

Deep-tech founders in Central and Eastern Europe are operating in a moment of structural change. Technology transfer models are evolving, geopolitical dynamics are shifting, and capital allocation patterns are redefining where and how companies scale.

For CEE founders, this creates both pressure and opportunity. The key differentiator will be how strategically you approach IP, ecosystem leverage, and global expansion.

‍

>>  Continuation. In Part I, we outlined the new deep-tech playbook for CEE founders — building strategic IP from day one, scaling capital-efficiently despite the growth gap, and turning the region’s talent and cost advantage into long-term competitive leverage.

‍

5. Protecting AI innovation: Patents vs. Trade secrets

The balance between patents and trade secrets to protect AI and data-driven innovations requires dynamic, TRL/stage-appropriate strategies rather than binary choices. AI poses unique challenges: algorithms evolve rapidly, datasets are non-patentable, and model training processes are difficult to detect when infringed. Special attention needs to be given to process/method patents, which are better kept a secret rather than published as a patent.

Framework for AI/deep-tech protection:

Patent the customer-facing innovation layer. File patents on: (a) novel system architectures and frameworks users interact with; (b) specific applications solving defined technical problems; (c) unique data processing or sensor fusion techniques. These provide enforceable rights where infringement is detectable through product observation.

Maintain trade secrets for competitive advantage elements: (a) training data and datasets, the proprietary information giving models their edge; (b) hyperparameters and optimization techniques, which can be the "secret sauce" of model performance; (c) model weights and internal architectures, when deployed server-side, where reverse engineering is infeasible.

Consider market dynamics and detectability. If deploying AI models on consumer devices or making their outputs public, patents may be necessary, as reverse engineering becomes feasible. For B2B SaaS or API-delivered services where models remain server-side, trade secrets often suffice.

Implement robust trade secret protection programs. Trade secret protection requires demonstrable efforts: (a) technical measures like access controls and encryption; (b) legal measures, including NDAs with all employees and partners; (c) physical security for sensitive systems. Without these, trade secret protection fails.

Geographic considerations matter. US trade secret law (DTSA) provides strong federal protection, but international enforcement is complex. For CEE startups targeting global markets, patents provide more reliable cross-border protection.

The Balkan opportunity: With 37% of Bulgarian STEM graduates being women (highest in Europe) and strong mathematical foundations across the region, Balkan teams excel at developing the algorithmic innovations suitable for trade secret protection. Their cost efficiency enables startups to maintain larger in-house development teams rather than outsourcing critical for protecting trade secrets.

Practical approach for early-stage startups: File provisional patents on core innovations to establish priority while maintaining trade secrets on implementation details. This preserves optionality - you can later decide whether to pursue full patents based on market traction and competitor activity, while trade secrets continue indefinitely if maintained properly.

‍

6. The next decade for Deep-tech in CEE: opportunities and threats

Major opportunities:

US retrenchment is creating an imperative for European sovereignty. The Trump administration's protectionist policies - 15% baseline tariffs, withdrawal from technology partnerships, and threats against EU tech firms - are forcing Europe to prioritize technological independence. This manifests as: (a) €100-135 billion potential AI-driven GDP growth in CEE specifically; (b) massive EU investment commitments including €640 billion in energy, €500 billion in general economy, and €35 billion in AI chips; (c) regulatory support through programs like CEE AI Action Plan, EIC funding, and regional AI sandboxes.

Deep-tech valuation premium. European deep-tech startups reach unicorn status in 6 years, compared to 8 for regular tech. With AI dominating - generative AI raised €56 billion globally in 2024 - CEE's technical talent is perfectly positioned. The region's 8 unicorns and 19 "soonicorns" demonstrate the pipeline is accelerating.

Balkan integration into the EU innovation architecture. Albania and Bosnia-Herzegovina joining the EIT Regional Innovation Scheme in 2025 creates formal pathways for Balkan innovators to access EU ecosystems, funding, and markets. Combined with Serbia's 51.5% innovation performance score and strong SME innovation rates, the Western Balkans are moving from an outsourcing hub to an innovation engine.

Quantum, nuclear fusion, and breakthroughs in advanced materials. Europe leads in several deep-tech domains - UK quantum investments (a recent £121 million commitment), nuclear fusion startups like Commonwealth Fusion and Seaborg, and advanced materials innovation. CEE universities' strong physics and chemistry programs position them to participate in these capital-intensive but potentially transformative sectors.

Defense tech’s emergence as a major vertical. With geopolitical tensions, defense tech attracted significant funding and investor interest in 2024-2025. CEE's proximity to security challenges and its integration into NATO create unique opportunities for dual-use technologies.

Major threats:

Growth-stage capital scarcity. While early-stage funding is robust (€1.7 billion seed funding in Europe Q3 2025), late-stage capital accounts for only 9% of global activity. Many CEE startups still relocate to the US for Series B+, creating brain drain. 2025 data shows CEE startup funding contracted to €510 million across 148 deals, concerning the given large deal dependence.

Talent retention challenges. Global competition for tech talent is intensifying, threatening CEE's cost advantage. As salaries rise, the region's primary competitive edge diminishes. Brain circulation programs are needed to reverse emigration.

Regulatory fragmentation. Despite EU membership, CEE faces implementation gaps. Romania, Bulgaria, and Croatia lag in R&D spending (all below 1% of GDP versus 3% EU target). Regulatory uncertainty around AI Act implementation, varying national innovation policies, and bureaucratic inefficiency slow growth.

US market access deterioration. Beyond tariffs, the Trump administration's aggressive stance toward EU tech regulation and withdrawal from collaborative frameworks makes US expansion costlier and riskier. Simultaneously, US venture capital's outperformance (€100 billion in AI investment versus the EU's €7 billion) continues attracting companies away.

Patent quality and enforcement gaps. CEE produces technical innovations but often lacks sophisticated IP commercialization expertise. Many universities and research institutes still treat IP as a cost center, and domestic patent filing rates lag Western Europe. Without strong IP portfolios, CEE startups struggle to compete for global capital.

China's overcapacity and pricing pressure. As US-China tensions redirect Chinese exports toward Europe, CEE companies face predatory pricing in clean tech, semiconductors, and other hardware-intensive sectors. This threatens emerging deep-tech hardware ventures.

‍

7. From Europe to the US: Navigating the most competitive tech market

The paradox of current US market conditions is that while Trump-era policies create barriers, they simultaneously weaken US startup competitiveness, creating openings for sophisticated European entrants. The administration's rollback of IRA clean tech subsidies, tariff-induced uncertainty, and regulatory attacks on EU companies are destabilizing capital allocation and talent retention in US tech hubs.

Five strategic imperatives for European deep-tech US entry:

1) Establish Delaware C-Corp structure early. This investor-friendly legal entity signals serious US ambitions and facilitates SAFE notes and stock option grants expected by American investors. Delay this decision and you create friction in every subsequent US negotiation.

2) Build authentic local presence through strategic hires. Success requires US-based leadership who understand local market dynamics - not European executives managing remotely. Bespoke Partners research shows effective US country leaders must integrate marketing, sales, delivery, and support functions while balancing HQ strengths with local capabilities. For deep-tech, this often means hiring a US Chief Revenue Officer or VP Sales with domain expertise and customer relationships.

3) Leverage US weakness in growth capital. While early-stage US funding dominates, many American deep-tech companies now struggle with Series B+ as VC consolidates around fewer mega-funds. European startups with strong IP portfolios, capital efficiency demonstrated through disciplined European scaling, and clear path to profitability can attract US growth capital seeking alternatives to overvalued Silicon Valley deals.

4) Target sectors where US advantages are eroding. The Trump administration's climate policy reversal creates opportunities in clean tech and industrial decarbonization - sectors where European regulation drives demand. Similarly, privacy-preserving AI and ethical AI become differentiators as US regulatory backlash builds. Enterprise B2B over consumer plays to European strengths in privacy, security, and regulatory compliance.

5) Use IP strategically for partnerships versus direct competition. Rather than competing head-to-head with established US players, European startups should structure IP licensing or joint ventures that give American companies access to proprietary technology while maintaining European operational control. This approach generated substantial revenue for companies unable to independently penetrate US markets.

Build multi-year US entry strategy, not rapid expansion. Data shows most successful European expansions take 18-24 months from initial market entry to meaningful revenue. Invest in understanding regional differences (US is not monolithic - Boston biotech ecosystem differs dramatically from Silicon Valley AI scene), customer buying cycles (US enterprise sales often involve longer technical validation), and cultural communication styles (Americans expect confident, benefit-focused pitching versus European tendency toward technical detail).

The current geopolitical window: US tariffs and protectionism, while challenging, may actually benefit European deep-tech long-term by keeping companies in Europe longer. Startups that previously would have relocated at Series A now scale in Europe through Series C, building more mature, capital-efficient operations before US expansion. This reduces dependence on US capital and creates stronger negotiating positions with American partners.