Thoughts on the Market
Episode: "Europe in the Global AI Race"
Host: Paul Walsh (Morgan Stanley's European Head of Research Product)
Date: November 13, 2025
Location: Live from Morgan Stanley’s 25th European TMT Conference in Barcelona
Panelists:
- Adam Wood (Head of European Technology and Payments)
- Emmett Kelly (Head of European Telco and Data Centers)
- Lee Simpson (Head of European Technology Hardware)
Overview: Europe's Position in the Global AI Race
This episode delves into Europe’s role and prospects in the global Artificial Intelligence (AI) race, particularly focusing on “agentic AI”—AI systems capable of autonomous action. Live from the 25th European TMT Conference, host Paul Walsh and fellow Morgan Stanley analysts explore how Europe can keep pace with the US and Asia in AI technology, adoption, and infrastructure, and what the implications are for the broader tech and investment landscapes.
Key Discussion Points and Insights
1. Framing the “Agentic AI” Opportunity in Europe
(00:56 – 05:31)
-
Automation Revolution:
- Adam Wood explains how generative and agentic AI are dramatically increasing the scope of automation, allowing tasks previously left to human labor (e.g., nuanced customer interactions or finance tasks) to be addressed by machines.
- "If we expand massively the capabilities of what software can automate, surely that has to be good for the software sector." (Adam Wood, 03:29)
- Past automation (“RPA”) had limits—could only reach about 30-40% of processes. With agentic AI, this ceiling is being broken.
- Adam Wood explains how generative and agentic AI are dramatically increasing the scope of automation, allowing tasks previously left to human labor (e.g., nuanced customer interactions or finance tasks) to be addressed by machines.
-
Call Center Example:
- AI can now predict or automate customer service responses, representing a paradigm shift from outsourcing/offshoring to total automation.
-
Enterprise Adoption Lag:
- While consumer AI adoption is visible, enterprises are lagging; Adam expects this to be a timing issue, not a signal of lack of opportunity.
2. Investor Pushback and Challenges in Software Adoption
(04:06 – 05:41)
- Adam addresses three main investor concerns regarding widespread enterprise AI adoption:
- Cheaper/Commoditized Coding:
- Ease of building software may reduce the barriers to entry, leading to price pressure from new entrants.
- "The barriers to entry of software have just come down dramatically… it's going to be new entrants that challenge the incumbents." (Adam Wood, 04:46)
- Ease of building software may reduce the barriers to entry, leading to price pressure from new entrants.
- In-house Development vs Vendor Reliance:
- Companies could start building their own solutions instead of using third-party platforms.
- Seat-based Pricing Models:
- As AI makes organizations more efficient, they may need fewer software “seats,” reducing vendor revenue even as enterprise efficiency increases.
- Cheaper/Commoditized Coding:
3. Semiconductor and Hardware Sector’s Role
(05:41 – 08:53)
-
Lee Simpson:
- The “hardware” piece—especially semiconductors and related equipment—has become a European “bonanza” as GPU and server architectures evolve.
- The US has historically led in this space, but Europe is now becoming more integral as global AI-driven demand expands.
- "We'd been seen as laggards... but now we're starting to see a flipping of that. These are beneficiaries." (Lee Simpson, 06:44)
-
Global Data Center Build-Out:
- Debate is less about “if” and more about “how much” infrastructure: can the world sustain massive, growing power demands (20–35 gigawatts/year by 2027–28)?
-
Investment Shifts:
- More global investors are now considering European hardware opportunities in this AI build-out, not just the traditional US and Asian plays.
4. Telco & Data Center Perspectives
(08:53 – 10:39)
- Emmett Kelly:
- Telcos remain largely on the sidelines of the AI wave; stock market flows have shifted away from telcos towards tech, especially in the US.
- "Telcos have struggled to put their case forward about how they can benefit from AI." (Emmett Kelly, 09:29)
- While telcos have discussed chatbots and smart networks, they haven’t significantly advanced their positioning.
- Data center investments are extremely capital intensive—rolling out a modern 100MW data center can cost ~€3 billion.
- Telcos are not heavily involved in large-scale data center builds; these are out of reach for all but the largest, most specialized players.
- Telcos remain largely on the sidelines of the AI wave; stock market flows have shifted away from telcos towards tech, especially in the US.
Notable Quotes & Memorable Moments
-
Adam Wood on Legacy Automation Limits (02:06):
"It was quite difficult with traditional coding techniques to go a lot further… we'd leave it with labor and do the labor arbitrage... it's practically impossible [to code for every call center response]." -
Lee Simpson on Europe’s Hardware Upside (06:44):
"We'd been seen as laggards... but now we're starting to see a flipping of that. These are beneficiaries." -
Emmett Kelly on the Telco Struggle (09:29):
"Telcos have struggled to put their case forward about how they can benefit from AI… they haven't really advanced their case since then."
Timestamps for Key Segments
- 00:56 – 03:29: Adam Wood discusses enterprise software's evolving role via agentic AI.
- 04:06 – 05:41: Pushback and challenges facing software vendors in the AI era.
- 05:41 – 06:44: Lee Simpson discusses semiconductors and Europe's role.
- 07:16 – 08:53: Debate on global data center build-out and European investment dynamics.
- 09:04 – 10:39: Emmett Kelly covers telcos’ relationship to AI and the scale/cost of data centers.
Conclusion
Europe stands at a crucial juncture in the global AI race. While previously characterized as a laggard, the continent is seeing new opportunities, especially in hardware and infrastructure. Agentic AI is expanding what software—and thus software businesses—can do, despite lingering concerns about commoditization, pricing, and enterprise adoption complexity. Infrastructure build-out, particularly for semiconductors and data centers, is both a challenge and an area of opportunity. Meanwhile, telcos struggle to stay relevant in this narrative. The wider debate in the coming years will focus not just on technology, but also on investment structures and competitive positioning between global regions.
Stay tuned for the follow-up episode, which will further dig into tech disruption and data center investments.
