The World Economic Forum published a report titled “The Future of Venture Capital: Unlocking Liquidity and Growth,” describing pressure on the traditional VC recycling model and a buildup of unrealized value in private portfolios. The WEF reports nearly $3.5 trillion in global venture capital assets under management and estimates $530 billion in global deal value for 2025. The report states an estimated $3.2 trillion of unrealized value is held in venture fund portfolios, including $1.7 trillion in funds launched in 2019 or earlier (World Economic Forum).

iConnections’ 2025-2026 allocator survey finds AI has shifted from a thematic allocation to a core portfolio lens. The iConnections research reports 54% of allocators invested in AI-related assets in 2025, rising to 83% of allocators naming the AI supercycle as a top theme by 2026. The survey also reports 44% of allocators see AI infrastructure as the most compelling opportunity (iConnections).

Industry-pattern observations: Reporting and allocator surveys together indicate capital demand is concentrating on the physical and digital backbone of AI, data centers, semiconductors, energy capacity and related supply chains. Allocators’ preference for infrastructure exposure aligns with documented increases in AI-driven capex and with public commentary that earlier-stage capital is being stretched by larger scale-up requirements.

The combination of large unrealized private-market value and a structural shift in allocator preferences toward AI-related infrastructure complicates traditional VC valuation and liquidity mechanics. For limited partners, longer fund timelines and locked-up distributions, as reported by the WEF, reduce the speed at which capital is redeployed. For allocators, the iConnections findings indicate a tilt toward exposures that capture AI buildout economics rather than only software or application-level winners.

Observers should monitor three indicators:

• •the pace at which secondary-market mechanisms or new fund structures surface in the WEF report recommendations
• •allocation shifts in large public and private LPs toward AI infrastructure categories (data centers, chips, energy)
• •evidence of realized exits or structured liquidity programs that reduce the reported $3.2 trillion unrealized backlog

Tracking these will show whether reported preferences translate into redeployed capital and altered valuation multiples.