AI startups absorbed $242 billion in venture capital in Q1 2026 alone. That’s 81% of all startup funding globally — a concentration that hasn’t been seen in any sector since the internet bubble.
In the same week those numbers landed, Elad Gil — one of the most prolific AI investors in Silicon Valley, with billions deployed across the sector — publicly advised AI founders to consider selling their companies within 12 to 18 months.
When the person writing the checks says the party might be ending, it’s worth thinking about what that means for your business.
The Boom by the Numbers
The scale of AI investment right now is genuinely unprecedented. Consider the data points from this week alone:
- $242 billion in VC funding flowed to AI startups in Q1 2026
- AI captured 81% of all global startup funding — crowding out nearly every other category
- DeepSeek is seeking a $20 billion valuation in its first fundraising round, with Tencent and Alibaba in discussions
- VAST Data closed its Series F at $30 billion, tripling its valuation in roughly two years
- SpaceX secured an option to acquire AI coding startup Cursor for $60 billion
These aren’t incremental numbers. They represent a market that’s pricing AI companies at levels where the gap between valuation and revenue is wider than it’s been since the late 1990s.
The money is real. The demand is real. But the question Gil and other experienced investors are quietly asking is: how long can this level of enthusiasm sustain itself?
Gil’s Argument: Why He Sees a Window Closing
Elad Gil isn’t a casual observer. He’s invested in some of the biggest AI companies of this cycle and has a track record of making early, contrarian calls that proved correct. His argument has several layers:
Competition is compressing differentiation windows. Two years ago, having a capable AI model was a moat. Today, OpenAI, Anthropic, Google, Meta, and dozens of well-funded startups all offer frontier-quality models. The technology layer is commoditizing faster than most founders anticipated. If your competitive advantage is “we use AI,” that advantage is evaporating.
The 1995-2001 parallel is uncomfortable but relevant. Gil explicitly draws the comparison to the internet boom. Not because AI isn’t real — the internet was real too — but because markets have a pattern of overpricing real technology. The companies that survived the dot-com crash were the ones with genuine revenue, real customers, and sustainable unit economics. Many of the most hyped companies didn’t.
Capital abundance creates a false signal. When capital is cheap and abundant, companies that should be struggling look healthy. High valuations mask weak business fundamentals. When capital tightens — and it always does eventually — the companies running on investor subsidies rather than customer revenue face a reckoning.
The window for favorable exits is finite. In a boom, acquirers pay premium prices and IPO markets are receptive. That doesn’t last forever. Founders who wait for the “perfect moment” often find themselves on the wrong side of a market correction.
What the Bears Get Right
Even if you’re optimistic about AI’s long-term trajectory, the bearish case has some grounded points:
Margin pressure is building. Compute costs remain high. Model training is expensive. The price war between major providers (OpenAI cutting prices, Anthropic competing, open-source alternatives proliferating) is squeezing margins for startups that are wrappers around foundation models. If your business depends on arbitraging the gap between model cost and customer willingness to pay, that gap is narrowing.
Customer fatigue is emerging. Enterprise buyers who enthusiastically signed pilot contracts in 2024 and 2025 are now asking harder questions about ROI. “We deployed AI” is no longer impressive; “We got measurable results from AI” is the new bar. Startups without clear, demonstrable value are going to face tougher renewal conversations.
The talent market is shifting. AI engineers are still in demand, but the extreme salary premiums are starting to normalize. This is good for the industry long-term but rough for startups whose pitch to employees was partly “join us and ride the wave.” As the wave’s momentum becomes less certain, retention gets harder.
Regulatory overhang is real. The EU AI Act is in effect. Australia, Canada, and multiple US states are advancing their own frameworks. Compliance costs will disproportionately impact smaller companies. Some startups will find that the regulatory burden makes their business model uneconomic.
What the Bulls Get Right
The bear case has merit, but it’s incomplete. The bull case for AI isn’t based on speculation — it’s based on revenue that’s actually flowing:
Enterprise demand is real, not speculative. OpenAI reportedly generates $2 billion in monthly revenue. Google Cloud’s AI services are growing rapidly. Anthropic, Databricks, and dozens of vertical AI companies have genuine enterprise customers paying real money. This isn’t 1999-era eyeball-counting.
AI is deflationary, not just generative. Unlike many tech waves that created new spending categories, AI is reducing costs in existing categories: customer support, software development, content creation, data analysis. Businesses adopt AI because it saves money, not because it’s trendy. That’s a more durable demand driver.
The infrastructure investment is self-reinforcing. Google just committed to new TPU generations. Amazon partnered with Anthropic for gigawatts of compute. Microsoft continues to expand Azure AI capacity. This infrastructure spending creates an ecosystem that sustains demand even through market corrections.
Productivity gains are compounding. Google reports 75% of new internal code is AI-generated. Companies using AI agents report significant efficiency improvements. These gains don’t reverse when market sentiment shifts — they’re structural.
Framework: How to Decide for Your Startup
Gil’s advice isn’t necessarily wrong, but it’s also not universal. The right response depends on your specific situation:
Assess your moat honestly. If your competitive advantage is primarily “we use AI” or “we’re a wrapper around GPT/Claude,” Gil’s warning applies directly to you. If you have proprietary data, deep domain expertise, strong customer relationships, or network effects, you’re in a different position.
Examine your revenue quality. Are your customers paying because your product delivers measurable value? Or are they paying because their boss said “we need an AI strategy”? Pilot-driven revenue and hype-driven revenue don’t survive market corrections. Outcome-driven revenue does.
Stress-test your unit economics. What happens if you can’t raise your next round? What happens if compute costs don’t decrease as fast as you projected? What happens if your largest customer doesn’t renew? If any of these scenarios breaks your business, you’re more fragile than your valuation suggests.
Consider your competitive timeline. How long until a larger player builds what you offer? If Google, Microsoft, or OpenAI can replicate your core functionality as a feature, your window is shorter than you think. If your value comes from something they can’t easily replicate — domain data, customer workflow integration, regulatory expertise — you have more time.
Evaluate your personal goals. This gets overlooked. If you’ve been building for five years and a favorable exit would be life-changing, the risk-reward math of “hold and hope for more” is different than if you’re 18 months in with a small team.
Practical Moves Beyond “Sell or Don’t”
The conversation doesn’t have to be binary. There are practical options between “sell everything” and “ignore the warning”:
Secondary sales. Founders and early employees can sell a portion of their equity on secondary markets, locking in some returns while maintaining full operational control. This is increasingly common and available at earlier stages than it used to be.
Strategic partnerships. Deeper partnerships with larger players can provide revenue stability, distribution, and a potential soft landing if the market shifts. Google’s $750M partner fund, announced today, is literally designed for this.
Diversified revenue streams. If you’re 100% dependent on one product or one type of customer, use the current market strength to diversify. Add professional services, build for adjacent markets, or layer in usage-based pricing that’s more resilient than contract-based ARR.
Defensive fundraising. If your fundamentals are strong, consider raising now even if you don’t need the capital immediately. Having 24-36 months of runway transforms a potential market downturn from existential to uncomfortable.
Controlled growth. Not every company needs to be a rocketship. If you can reach profitability or near-profitability at your current scale, you become largely immune to fundraising market conditions. Cash-flow positive is the ultimate hedge.
The Bottom Line
Elad Gil isn’t predicting a crash. He’s saying the current window of extraordinary valuations and easy capital won’t last forever — because it never does. That’s not a controversial claim; it’s a historical observation.
The founders who should take this most seriously are those whose businesses depend on continued market euphoria rather than customer value. If your company needs favorable fundraising conditions to survive, the clock is ticking.
The founders who can afford to be calmer are those with real revenue, genuine differentiation, and customers who would miss them if they disappeared. These businesses will thrive regardless of what the VC market does.
The smart move isn’t panic. It’s honest assessment. Stress-test your position. Know your options. Make deliberate choices while the market is still in your favor — because waiting until it shifts means having fewer options and less leverage.
$242 billion in a single quarter says the boom is real. An experienced investor warning about the cycle says the boom is finite. Both things can be true at the same time. Your job as a founder is to act on both.
—
Next Steps
Building in AI and wondering about your strategic positioning? OpenVerb helps founders stress-test their market position and plan ahead. [Get in touch](https://openverb.com/contact) for a strategy conversation.