At Rise of AI 2024 in Berlin, Sabine Wiesmüller, Director of the AI Competence Center at Start2 Group, argues that European AI startups need to define global niches and expand into other regions rather than trying to mimic the United States. She outlines Europe's structural weaknesses: relatively low global AI market share, weak research-to-application transfer, startups that stay too small for one region, roughly 75% of European exits going to non-European buyers, and a persistent gap in late-stage investment. She then takes the audience through opportunity hubs around the world, including Boston and Sao Paulo in the Americas, Singapore, Beijing and Bangalore in Asia, and manufacturing and fintech growth in Chile and Brazil, citing statistics on elite AI talent shifting from the US toward India and China between 2019 and roughly 2022. She closes with market trends such as agentic AI, specialized large language model applications, and AI-driven drug discovery, and advises startups and corporates to treat ecosystems, regulatory differences and AI adoption strategically rather than defensively.
Key takeaways
- Roughly 75% of European startup exits go to buyers outside Europe, which Wiesmüller ties to a persistent gap in late-stage investment.
- The share of elite AI researchers based in the US fell from about 60% to 42% between 2019 and roughly 2022, with China rising to 28% and Europe at 12%.
- German SMEs remain hesitant to pool even anonymized data, limiting their ability to train models like predictive maintenance systems that need negative cases.
- Foxconn invested $700 million in AI chip manufacturing in Jalisco, Mexico, partly to reduce dependence on China.
- Brazilian manufacturers using AI have cut average downtime by 25%, and Latin America's AI market is projected to grow 28.5% to $10 billion by 2030.
Detailed notes
Europe's structural gaps
Wiesmüller argues that Europe needs to take the driver's seat in AI rather than mimic the US, and identifies where Start2 Group, which helps startups scale globally through legal entities in Europe, the US and Asia, sees Europe falling short. She says Europe's global AI market share is limited partly because research is not transferred into application effectively, and once startups do mature, they often fail to expand beyond one regional market. She cites that roughly 75% of European startup exits go to buyers outside Europe, a trend she says is tied to a persistent gap in late-stage investment, since venture capital remains scarce after COVID and most VCs focus on their existing portfolios rather than new bets.
Regional opportunities in the Americas
Wiesmüller highlights Boston as the leading US biotech hub, noting the FDA currently has shorter approval times for life science solutions than the EU. She contrasts Silicon Valley's heavy investment and close ecosystem collaboration, including data pooling, with what she describes as more siloed thinking in Germany, particularly among SMEs reluctant to share even anonymized data needed to train predictive maintenance models on negative cases. She names Sao Paulo as a major fintech hub she says surprised her, with banks and governments among the clients, climate-affected regions as growing markets for clean-tech AI solutions, and Chile, where a client construction firm shows a Silicon Valley-oriented mindset comparable in maturity to European SMEs.
Asia: Singapore, Beijing and Bangalore
Wiesmüller describes Singapore as one of the region's largest AI investors, with expedited patent processes supporting faster application of research, and cites Nvidia's heavy investment across Asia, including plans for a supercomputer in Japan and India. She names Beijing for its focus on facial recognition and deep learning, noting China has broad application and large datasets but not always high data quality. She highlights Bangalore, sometimes called India's Silicon Valley, for strong ecosystem-wide data sharing and Reliance Jio's partnership with Nvidia.
Talent migration and funding trends
Wiesmüller presents figures showing the share of top 2%-tier AI researchers originating from the US fell from 35% in 2019 to 28% three years later, converging with India and China, while the share of such researchers working in the US fell from about 60% to 42% over the same period, with China at 28% and Europe at 12%. She notes that Chinese firms such as Baidu, Alibaba and ByteDance released large language models the prior year, creating talent pools startups can leverage. On funding, she says US VCs increasingly favor existing portfolio companies, and that in Latin America about 47% of investment goes to Brazil, mostly Sao Paulo, while Foxconn's $700 million investment in Jalisco, Mexico, in AI chip manufacturing illustrates a push for supply-chain independence from China.
Market trends and closing advice
Wiesmüller points to agentic AI and multimodal AI as cutting-edge US trends, and argues large language models still offer opportunity through specialization for specific industries and regulations, contrary to the impression in Germany that the space is already saturated. She cites AI-driven drug discovery with shorter US approval times, Brazilian fintechs such as Nubank applying AI more aggressively than European banks, and Brazilian manufacturers cutting downtime by 25% on average through AI, with Latin America's AI market projected to grow 28.5% to $10 billion by 2030. She closes with three takeaways: seek out ecosystems that offer money, talent, knowledge and clients; recognize that markets differ in maturity, regulation and client needs, for example Brazil's more permissive fintech nudging rules; and treat AI adoption as essential for every company in a data-driven economy.
Questions answered in this talk
Why do so many European AI startups fail to scale globally?
According to Sabine Wiesmüller at Rise of AI 2024, European startups often stay tied to one regional client base instead of expanding market to market, and about 75% of exits go to non-European buyers because of a persistent gap in late-stage investment.
How has the geography of top AI talent changed in recent years?
Wiesmüller cites data showing the share of the world's top 2% AI researchers working in the US fell from about 60% in 2019 to 42% roughly three years later, with China rising to 28% and Europe at 12%, alongside a similar shift in researchers' countries of origin.
Which emerging markets does Wiesmüller highlight for AI startups?
She points to Boston and Sao Paulo in the Americas, Singapore, Beijing and Bangalore in Asia, and Chile and Brazil for manufacturing and fintech, noting strengths such as Singapore's expedited patent processes and Brazil's fast-growing fintech sector.