Anthropic’s run-rate revenue from Europe, the Middle East and Africa has surged more than ninefold year-on-year, the company has confirmed, as it prepares to open its sixth European office in less than twelve months. The Milan outpost, due to launch this month, marks a further acceleration of the AI firm’s international push, with the region now described as its fastest-growing geography.
The Milan office follows openings in London, Dublin, Zurich, Paris and Munich, and comes as Anthropic plans to triple its international workforce to keep pace with demand for its Claude AI models. The London office alone employs around 200 staff and is expected to expand to 800. To attract the necessary talent, the company is offering London engineers salaries of up to £630,000 per year before stock options – a compensation level that, according to industry observers, is already forcing European startups to reconsider their own pay structures.
The number of large business accounts in EMEA has grown more than tenfold over the same period, driven largely by enterprise adoption. “After France and Germany, Italy is a natural next step,” Chris Ciauri, Anthropic’s managing director of international, told Italian newspaper Corriere della Sera.
The Vatican and the governance play
Anthropic’s move into Italy is not purely commercial. Co-founder Christopher Olah appeared alongside Pope Leo XIV on 25 May at the presentation of the Vatican’s first major teaching document on AI ethics, the encyclical Magnifica Humanitas, speaking in the Synod Hall before cardinals, theologians and the diplomatic corps. CEO Dario Amodei then travelled to Rome for meetings with senior Italian institutional figures during the same week – a sequence that positions the company at the heart of the global AI governance conversation.
That strategy is especially pointed given the scrutiny Anthropic has faced in the United States. The Pentagon labelled the company a “supply chain risk”, the first time such a designation has been applied to an American firm. The dispute, according to sources familiar with the matter, centred on contract terms related to mass domestic surveillance and autonomous weapons systems. Anthropic has said it plans to sue the government over the label. In Rome, by contrast, it is embracing the role of governance partner, leaning into the moral and regulatory authority of the Vatican.
Why Milan matters commercially
Italy gives Anthropic access to a Southern European industrial economy built around financial services, manufacturing and consumer goods – precisely the sectors that are most sensitive to the EU’s emerging regulatory framework. The EU AI Act is being implemented in stages: prohibitions on certain AI practices and AI literacy obligations took effect in February 2025; transparency rules for general-purpose AI models followed in August 2025; and requirements for high-risk AI systems become applicable from August 2026, with full applicability by August 2027. European companies, particularly in Italy, are acutely aware that US frontier AI models typically arrive with compliance frameworks calibrated for American legal standards, not EU law.
Anthropic is betting that its existing governance credentials – cemented by the Vatican engagement, the company’s safety-focused brand and its early compliance work – will give it an edge. The firm already works with European enterprises including L’Oréal, BMW, SAP and N26, and has been embedding Claude into enterprise operations through a $1.5bn joint venture with Goldman Sachs and Blackstone. Milan extends that footprint into a market that offices in Paris and Munich do not naturally serve from a language, regulatory or relationship standpoint.
The competitive dynamics are clear. OpenAI, Anthropic’s main rival, does not have a dedicated office in Italy. Across Europe the two are racing to cement their presence: OpenAI had opened six European offices by the end of 2025, with a seventh in Munich, and is also expanding in London. But market share data from Menlo Ventures’ end-of-2025 State of Generative AI report shows Anthropic’s share of the enterprise large-language-model market has risen to approximately 40%, up from 32% in mid-2025 and just 12% in 2023. OpenAI’s share has fallen over the same period from 50% to 27%. Anthropic now reports that the number of enterprise customers spending $1m or more annually on Claude doubled in under two months to surpass 1,000.
Anthropic has also secured significant computing resources to underpin its growth, including a nearly $45bn deal with SpaceX and a $1.8bn agreement with Akamai Technologies, alongside its existing use of Google’s chips and cloud services. A separate deal with Microsoft will integrate Claude into the Copilot assistant.
The financial trajectory underscores the pace. Anthropic crossed a $30bn annualised revenue run rate in April 2026, having been at roughly $9bn at the end of 2025 and $1bn in January 2025. The company closed a $30bn Series G funding round in February 2026 at a $380bn valuation, and is reportedly in talks for a new raise that could value it at over $900bn – more than doubling in roughly three months. CEO Dario Amodei has stated that Q1 2026 saw 80x annualised revenue growth, outstripping the company’s own forecasts by a factor of eight, though those figures are CEO-attributed and have not been independently verified through audited financials. Anthropic is also said to be on pace for its first profitable quarter.
Europe’s most valuable homegrown AI company, Mistral AI, has been expanding aggressively in parallel – raising €1.7bn and breaking ground on a dedicated AI data centre south of Paris – but has no Italian office of its own. The question Europe’s AI ecosystem now faces is whether the arrival of well-resourced, safety-branded US incumbents accelerates the continent’s adoption or crowds out the European alternatives trying to build their own.
