The 2025 Lux Innovation Survey: How AI and Deglobalization Are Rewriting Corporate Innovation 

Recorded by:

Written by:

Innovators are moving fast on AI, even as strategy, governance, and KPIs trail deployment. In parallel, leaders expect global markets to persist but are actively regionalizing parts of their innovation approach to manage risk and shifting demand. Uncertainty is the dominant theme. The mandate for 2025: Tighten AI strategy before scaling and redesign operating models for a more regional yet still globally connected innovation system.  

Survey participants 

The survey reached roughly 100 senior innovation leaders across the Americas, EMEA, and APAC, most with more than 15 years of experience in R&D and innovation management. Many oversee innovation portfolio management, R&D management, and corporate strategy.  

AI in innovation: confidence is high, maturity is mixed 

Leaders expect AI to accelerate speed and agility across the innovation lifecycle, especially for document review, literature search, knowledge management, competitive intelligence, and R&D productivity. Investment decision support lags in confidence compared to other areas.  

Yet, governance and measurement are not keeping pace: 

  • About one-third reported having an AI strategy. Only 13% have a dedicated AI executive, and just 7% have AI-related KPIs. Roughly one-quarter report having a dedicated AI budget.  
  • Despite this, about half say they have already deployed AI or will do so this year. Budget expectations over the next three years are scattered.  

Why this matters: Organizations risk fragmented, disappointing implementations if deployment outruns intent, governance, and metrics. Lux experts advise setting strategy and desired outcomes first, even under executive pressure to move quickly.  

What could slow AI? 

Respondents flagged security, skill loss, and bad data as leading risks. Security concerns center on proprietary data exposure and vendor IP terms; skill loss reflects anxiety about over-reliance on AI and uniform thinking.  

Practical implication: Most innovation use-cases do not require model training, which can reduce some security risks. Companies can further mitigate exposure by avoiding vendors with unfavorable IP terms or hosting models locally.  

Deglobalization and trade: Global in principle, regional in practice 

Leaders remain broadly confident that the economy will function as a global marketplace. About 80% agree it will stay global. At the same time, 34% agree that multiple, largely independent economic zones will emerge, and 40% expect market demands to diverge.  

Here’s what’s changing: 

  • A shift from one global playbook to more regionally balanced strategies over the next three years is changing company strategies.  
  • A stronger move toward balanced, globally coordinated team composition and broader participation in innovation networks is changing company teams and networks:.  

Why the shift? Even if core commodity markets remain global, product needs, regulations, and cost structures are diverging by region. Leaders want more regional specificity in strategy with more global coordination in talent and networks to manage that complexity.  

Current innovation challenges 

Uncertainty ranks as the top challenge in responding to deglobalization. Many organizations are delaying decisions, with 65% of those without plans citing this reason. Coordinating across regions and time zones, managing rising costs, and navigating corporate bureaucracy also loom large.  

What to do next: A 90-day plan for innovation leaders 

  1. Codify AI purpose and guardrails 
    • Write a short, business-outcome-focused AI strategy that defines priority use- cases, expected value, and risk boundaries. 
    • Stand up a lightweight decision council and appoint a clear AI owner. Track a minimal set of KPIs such as cycle time reduction in literature review, percent automation in secondary research, and win rates for portfolio decisions enhanced by AI.  
    1. Sequence AI deployment to avoid fragmentation 
      • Start with high-signal, low-risk use-cases like knowledge management and competitive intelligence, then expand to R&D productivity tools. 
      • Choose vendors with acceptable IP terms or deploy models in controlled environments to reduce data leakage risk.  
        1. Rebalance your operating model for regional reality 
          • Move from a single global strategy to a regionally balanced one that still shares platforms, data, and governance. 
          • Broaden participation in cross-regional innovation networks to surface divergent needs early.  
            1. Accelerate decisions amid uncertainty 
              • Create a deglobalization “control tower” to track policy shifts, tariff impacts, and critical-materials constraints and trigger predefined responses. 
              • Delegate more decision rights to regional teams or streamline top-down approvals to improve speed. 

                What do you want to research today?