September 21, 2026, Hg announced an expansion of its strategic cooperation with Anthropic, with Hg Catalyst and its internal AI team working directly with Anthropic to deploy Claude’s Agentic AI capabilities to portfolio companies. The most significant incremental benefit for AI‑native acquisitions is that Hg links GP‑level AI capabilities, cutting‑edge models, and the portfolio companies’ system records and workflows into a repeatable post‑investment mechanism.
Facts
- Hg disclosed that the current portfolio contains more than 1,600 AI projects in operation, corresponding to a budget EBITDA impact of over USD 260,000,000; this metric is 5 times higher than at the start of the collaboration with Anthropic.
- The relationship began in early 2024, and the collaboration has already included joint AI hackathons, employee capability building, and embedded deployments; the newly added “strategic venture” pushes the cooperation further toward portfolio‑level Agentic AI deployment.
- Hg Catalyst is one of Hg’s AI product incubation and post‑investment value‑creation capabilities. Hg says its portfolio companies typically own industry‑specific system records and core workflows, while Anthropic provides the intelligence layer; the combination is seen as the key to generating operating value from Agentic AI.
- Hg cites Quantios as an example: Quantios, with support from Hg Catalyst, launched its first Agentic AI product in March 2026, after which Hg sold the company to Vista Equity Partners; Hg also mentioned recent exit cases involving GTreasury and Intelerad, noting that companies with trustworthy product‑embedded AI strategies are attracting stronger buyer interest.
Analysis
This update further standardizes Hg's post‑investment AI model: the GP retains AI product and engineering capabilities, frontier model companies provide the intelligence layer, and portfolio companies supply industry system records, data and workflows. It corresponds to a clear value‑creation chain—identifying reconfigurable workflows, embedding Agentic AI, creating replicable product and operating capabilities, and then applying the validated methods to other portfolio companies. For AI‑native buyouts, the metrics worth tracking have shifted from “how many tools are deployed” to “how many core workflows enter an Agentic closed loop, whether the budgeted EBITDA impact can be realized, and whether these capabilities improve exit quality and buyer demand.”
Counterevidence
- The USD 260,000,000 disclosed by Hg falls under “budgeted EBITDA impact,” representing a budgeted or projected effect and should not be regarded as realized or audited EBITDA increment.
- Hg links exits such as Quantios, GTreasury and Intelerad to its product‑embedded AI strategy, but public information does not demonstrate that AI is the sole determinant of transaction valuation or exit success.
- The parties have been cooperating continuously since 2024, and the incremental element of this event is mainly the upgrade of the collaboration into a more systematic strategic deployment mechanism; therefore, ongoing observation is needed to see whether verifiable increments appear in actual revenue, productivity, costs and exit multiples.
Financial News｜Hg agrees strategic venture with Anthropic｜2026-09-22