Carta's latest research confirms what family offices already suspected: 10% of new venture funds raised in each recent vintage hold commitments above $100 million, a ratio that has held without deviation since 2017. The 2017-2018 vintage cohort now tracks to strong LP returns, rewarding allocators who paid the premium for proven managers. The same week Carta published this data, OpenAI and Anthropic each announced private-equity-backed consulting ventures totaling $11.5 billion in aggregate capital, making explicit what top-tier fund managers have known for eighteen months—venture deployment now competes directly with venture capital for institutional attention.

The 10% figure represents persistence, not stagnation. Across the 2019, 2020, 2021, 2022, and 2023 vintages, the share of funds clearing $100 million in commitments remained flat while the denominator expanded sharply. Thousands of new managers entered the market during the zero-rate era, but allocator capital remained anchored to funds with multi-decade track records and structured co-investment rights. The consistency of this concentration suggests institutional LPs treat venture capital as a reputation-gated asset class where new entrants face a longer credentialing period than in any other alternative category. Firms that raised their first funds in 2020 or 2021 now approach their second closes with cap tables showing sub-$50 million commitments and limited follow-on interest.

What makes this week's datapoint relevant is the parallel capital raise in AI deployment infrastructure. OpenAI announced a $10 billion private-equity joint venture called the OpenAI Deployment Company, designed to embed AI systems into enterprise operations through consultancy work rather than software licensing. Anthropic revealed a complementary structure within 24 hours. Both vehicles bypass the venture fund model entirely, raising committed capital from institutional LPs who would otherwise allocate to top-decile venture managers. The overlap in investor profiles is direct: sovereign wealth funds, university endowments, and large family offices capable of writing $100 million+ checks now face a choice between backing proven venture franchises and funding the consultancy arms of the AI platforms those venture funds previously financed. For allocators, this creates a structural question about whether venture capital itself remains the optimal exposure to AI-driven returns, or whether deployment infrastructure offers a more certain path to realized gains.

The second-order effect is already visible in fundraising timelines. Several established venture firms with strong 2017-2018 vintage performance have extended their current fundraises by six to nine months, a schedule slip attributable not to LP skepticism but to bandwidth constraints as the same institutional allocators evaluate deployment-infrastructure opportunities. The 10% concentration ratio may hold in nominal terms, but the competitive set for that capital has widened beyond traditional venture structures for the first time since the asset class matured in the 1980s. Firms that relied on FOMO-driven closes during 2020-2021 now face an environment where institutional LPs perform multi-month diligence and request detailed portfolio-company integration plans before committing.

Allocators should watch for second-quarter fundraising data from PitchBook and Preqin, expected in mid-July, which will show whether the 10% ratio compressed in Q1 2026 as AI deployment vehicles absorbed capital that would have otherwise flowed to traditional venture structures. Separately, monitor whether the OpenAI and Anthropic consultancy arms report their first customer deployments before the end of Q2, as early traction will determine whether institutional LPs treat these vehicles as venture substitutes or as complementary exposures.

The 10% ratio persists, but for the first time in forty years, the asset class it describes now competes with a deployment model built by the same companies venture capital once funded exclusively.