监管障碍、资金短缺和劳动力技能缺口正促使欧洲初创企业考虑迁移，这可能引发该地区最具潜力的新创企业大规模外流，亚马逊云科技（Amazon Web Services）一位高管周三对路透社表示。
Regulatory hurdles, funding shortages and gaps in workforce skillsets are prompting European startups to consider relocating, risking a mass exit of the region's most promising new ventures, an Amazon Web Services executive told Reuters on Wednesday.
由于监管更严格、能源受限以及对数据隐私的更广泛担忧，欧洲在缩小与美国和中国在人工智能竞赛中的差距方面举步维艰。该地区鲜有自主研发的前沿人工智能模型，反而依赖OpenAI和Anthropic等公司。
Europe has struggled to narrow the gap with the US and China in the AI race due to tighter regulations, energy constraints and broader concerns about data privacy. The region has produced few frontier AI models of its own and is instead dependent on the likes of OpenAI and Anthropic.
“欧洲企业告诉我们，它们每赚1欧元可能要花费高达40美分用于合规，而非技术投入。想象一下，如果这笔钱重新投入技术研发会怎样，”亚马逊云科技欧洲、中东和非洲地区（EMEA）董事总经理Tanuja Randery在接受路透社全球市场论坛采访时表示。
"European businesses are telling us that they could spend as much as 40 cents to every euro on regulatory compliance instead of technology. Imagine if that was reinvested in technology," AWS EMEA Managing Director Tanuja Randery said in an interview with the Reuters Global Markets Forum.
她援引亚马逊云科技的一项企业调查称，近四成欧洲初创企业会考虑离开该地区以实现规模化发展。“在增长最快的企业中，这一比例超过一半。”
Nearly 4 in 10 European startups would consider leaving the region to scale, she said, citing an AWS survey of companies. "Among the fastest-growing, it’s over half."
“初创企业告诉我们，它们正在重新考虑选址，因为它们需要资金和支持，而在这里得不到足够的资金。”Randery表示，虽然欧洲超过一半的企业正在使用人工智能，但只有五分之一的企业明确的人工智能战略、拥有人工智能治理模型，并投资于能从根本上重塑运营的流程。她补充道：“我认为推理是欧洲企业真正需要去投资的领域。”
"Startups say to us that they're reconsidering locations because they want funding and support, and they're not getting enough funding here." While more than half the businesses in Europe are using AI, Randery said only one in five are clear about AI strategy, had an AI governance model and were invested in processes that would fundamentally reshape operations. "I think inferencing is where European businesses need to really go invest," she added.
人工智能推理是运行训练好的人工智能模型以生成输出的过程，也是模型部署用于现实世界应用的阶段。
AI inference is the process of running trained AI models to generate outputs and is the stage at which models are deployed for real-world use.
科技巨头正投入数千亿美元用于前沿研究和扩大人工智能基础设施，而欧洲的努力则受限于欧盟《人工智能法案》（AI Act），后者被广泛视为全球最严格的人工智能监管框架。严重的能源和电网限制也在阻碍进展。
While tech majors are pouring hundreds of billions of dollars into frontier research and scaling AI infrastructure, Europe's efforts have been limited by the European Union's AI Act, widely seen as the world's most stringent AI regulatory framework. Severe energy and grid constraints are also stymieing progress.
“我们需要在欧洲加快步伐……这不是进步与安全的对立。这是‘兼顾’，而非‘二选一’。”
"We need to go faster in Europe... It isn't about progress versus safety.
（加入GMF，这是一个托管在LSEG Messenger上的聊天室，用于进行实时访谈：
It's an 'and,' not an 'or.'" (Join GMF, a chat room hosted on LSEG Messenger for live interviews: