What do top VCs think will happen to AI valuations in the next 2 years?

Venture capitalists anticipate a bifurcated and volatile AI valuation landscape over the next two years, characterized by a ceiling-less market for top-tier companies alongside growing concerns of a speculative bubble. For elite AI startups, investors note there is currently no effective cap on valuations, fundamentally altering risk calculations for late-stage funds . This exuberance is fueled by unprecedented customer demand that, some argue, justifies the high prices , with the top three LLM companies alone projected to raise a combined $100 billion in a six-month period [21, 28]. Proponents view the valuations of leading foundation model companies as reasonable when analyzed through the lens of their high growth and revenue multiples . However, this optimism is tempered by significant skepticism, with some investors defining the current environment as a bubble, pointing to startups with zero revenue achieving valuations between $10 billion and $30 billion . This view is echoed by others who question how the hundreds of billions being invested will generate a financial return and draw direct comparisons between the current AI capital expenditure cycle and the telecom bubble that ended badly .

The debate over a potential bubble reveals a deep divide in long-term outlooks. Bullish investors believe the AI wave is an order of magnitude larger than previous technological shifts, predicting the creation of numerous trillion-dollar businesses and suggesting current high valuations still offer significant upside [9, 19]. This perspective is supported by the argument that demand for AI far exceeds supply, distinguishing the current market from past speculative manias . Conversely, prominent critics characterize the boom as a real technological wave inevitably paired with a speculative "industrial bubble," similar to the dot-com era . These concerns are amplified by the proliferation of "circular deals," where major tech companies invest in AI startups that then use the capital to purchase the investors' cloud or chip services, potentially inflating valuations and obscuring financial transparency . This sentiment is reflected in broader market surveys, where a majority of respondents identified a plunge in tech valuations and waning AI enthusiasm as the biggest risk to market stability in 2026 .

The disruptive force of AI is also expected to dramatically reshape valuations in the adjacent enterprise software market, though the precise impact remains contested. A bearish outlook suggests that AI will erode the value of traditional SaaS companies by rendering per-seat pricing models obsolete, causing investor uncertainty and capital flight from the sector [2, 5, 29]. Some predict this could cause SaaS price-to-sales multiples to collapse from 5-10x to as low as 1-2x, similar to utility companies . However, a counterargument from Andreessen Horowitz posits that the "Sasspocalypse" narrative is incorrect, noting that 75% of public SaaS companies have successfully raised prices since ChatGPT's release . This view holds that AI will primarily create new, native software categories and reduce switching costs rather than wholesale replacing existing tools . In this more optimistic scenario, vertical AI SaaS companies could become ten times larger than their predecessors by capturing operational spend, not just software budgets , with the "intelligence layer" of the AI stack capturing as much as half of the total enterprise value created .

Given this uncertainty, VC investment strategies are adapting to the new environment. Firms are remaining highly flexible on valuation for top-tier companies but are firm on ownership targets to align with their support models . The nature of market subsidy has also shifted from advertising spend to compute credits, viewed as a healthier form of customer acquisition . The most viable strategy for investors may be to back startups applying AI to niche verticals with proprietary datasets and complex workflows, as these are more defensible against large model providers . The short-term outlook is murky, with some analysts expecting clarity on AI's impact on SaaS within the next three to nine months , while others predict a market correction in specific sub-sectors like AI security within the next year . A significant near-term risk is the massive AI data center infrastructure build-out by hyperscalers, which could result in an oversupply reminiscent of the dot-com bust .