Anthropic and Hg plan a European AI consultancy with hubs in London and Berlin

Anthropic and Hg plan a European AI consultancy

Anthropic has agreed to back a new consultancy in Europe with Hg, the London private equity firm that specialises in software companies, according to a Financial Times report published on Friday 23 October 2026. The FT describes it as the AI lab's first move to create an independent firm selling AI expertise to companies on the continent. Neither company has announced the venture. Both declined to comment to the FT, so the details below come from people familiar with the plans rather than from Anthropic or Hg.

The new company would place AI consultants and engineers inside client businesses to help them design and deploy products built on AI, according to The Next Web's summary of the report. It is expected to launch by the end of 2026, with hubs in London and Berlin. The timing matters: Anthropic is preparing a stock market listing and its draft IPO filing showed both very fast revenue growth and very large losses, so investors are watching every new source of income.

What the new firm would do

The consultancy's main targets, according to the people cited by the FT, are software companies that sell to other businesses, along with other white-collar industries. Its engineers would either lead the design of new AI-powered products at a client or support AI projects that are already under way. The firm would also take on work across companies that Hg itself owns.

That is the model the industry calls forward deployed engineering: instead of selling access to a model and leaving the customer to work out the rest, the vendor sends engineers who sit with the customer's team until something useful ships. It is labour-intensive and hard to automate. A recent benchmark found that the best AI agent resolved only 14 of 95 customer delivery tasks, which helps explain why labs still want humans on site for this work.

Hg already employs about 150 AI specialists who help introduce the technology across its portfolio. One person told the FT that the new consultancy would operate alongside that team rather than replace it, so Hg companies that rely on the in-house group today would keep the same support.

How the money and the people would work

On the money side, Anthropic would fund its share of the venture from its own balance sheet, while Hg would invest through the funds it manages, including the London-listed HgCapital Trust, according to the report. That last detail matters for public investors, because it means a listed investment trust would hold part of the new firm. The reports do not say how much either side is putting in, how the equity is split, or who will run the company.

The two companies are not starting from zero. On 21 September 2026, Hg announced a new strategic venture with Anthropic to bring Claude's agentic capabilities to its portfolio, marking three years of working together. In that announcement Hg said its relationship with Anthropic dated back to early 2024, that it had more than 1,600 live AI projects across its portfolio, and that those projects represented over $260 million of budgeted EBITDA impact. The two firms have run a joint AI hackathon series, staff training and embedded deployment work.

The September announcement was about Hg's own AI team, Hg Catalyst, working directly with Anthropic. The new consultancy, as described by the FT, would go further: a separate business with its own staff that can sell its services to companies outside the Hg portfolio as well.

A template Anthropic used in the US first

Anthropic has done this before in the United States. In May 2026, Anthropic, Blackstone, Hellman & Friedman and Goldman Sachs announced a new AI services firm that would work with companies to bring Claude into their core operations. Anthropic's chief financial officer, Krishna Rao, said at the time that "enterprise demand for Claude is significantly outpacing any single delivery model." Reports have put the planned investment at $1.5 billion.

In July that firm launched under the name Ode with Anthropic. It is built on Fractional AI, an applied AI services firm acquired in May, whose co-founders Chris Taylor and Eddie Siegel became chief executive and chief technology officer. Its investor group includes General Atlantic, Leonard Green & Partners, Apollo Global Management, GIC and Sequoia Capital. Anthropic has said Ode is aimed at mid-size companies moving from experimenting with AI to building it into their operations.

OpenAI has followed the same path. In May it launched the OpenAI Deployment Company with the private equity firms TPG, Advent, Bain Capital and Brookfield, according to the reports. The pattern is now clear: AI labs pair their models with private equity money and with engineers on the ground, and the private equity firms get a ready channel into the hundreds of companies they own.

Why Hg, and why now

For Hg, the logic is partly defensive. Software stocks sold off earlier this year on fears that AI would undercut traditional business software, and private equity owners have been under pressure to show that the companies they bought over the past decade can become AI businesses rather than AI casualties. Hg has still found buyers, selling Gen II to KKR for more than $5 billion and Quantios to Vista Equity Partners for about 400 million pounds, according to SQ Magazine.

For Anthropic, a European services arm offers revenue and reach. The FT reported last week that Anthropic's revenue growth slowed in the third quarter, a quarter in which it told investors revenue reached $16.8 billion. Services work is lower margin than selling model access, but it pulls more of a customer's AI spending towards Claude and creates engineers who know how to make Claude work inside specific industries.

What it means for consultancies and customers

These ventures put AI labs in more direct competition with established IT consultancies such as Accenture and the Big Four accounting firms, Deloitte, EY, KPMG and PwC, for whom advising companies on technology adoption is a large business. The FT noted that tension. The big consultancies are also partners of the labs, and some AI companies have signed alliances with them, as Cohere did in its enterprise AI alliance with PwC.

For European companies, especially software makers in Hg's portfolio, the practical effect would be access to engineers who work closely with Anthropic and see its newest models early. The risk for customers is lock-in. A consultancy part-owned by a model maker has an obvious incentive to build on that maker's models, and buyers will want to know how neutral its advice can be when a different model fits a job better.

What is still unknown

The biggest gaps are the ones the reports leave open: how much each partner is investing, who owns what share, who will lead the firm, how many engineers it will hire, and whether it will use Claude exclusively. None of that has been confirmed, and the plan could still change before launch.

The next signals to watch are an official announcement from Anthropic or Hg, any disclosure from HgCapital Trust about its commitment, and Anthropic's public IPO filing, which may describe its services ventures and how it accounts for them. Until then, this is a well-sourced report of a plan, not a launched company.