Elena Volotovskaya is the Head of Softline Venture Partners.
AI projects have become the axis of capital flows and the main focus for investors. In 2025, AI startups raised $192.7 billion in global venture capital.
At the same time, while some projects succeeded (or at least managed to stay afloat), others failed to turn their ideas into sustainable businesses. Here, I’d like to examine the reasons behind AI startup failures and how venture investors should reconsider their approach to funding AI ventures in 2026.
Why Do AI Startups Fail?
“There’s a little bit of a hype bubble going on in the early-stage venture space,” Reuters quotes Bryan Yeo, chief investment officer at GIC Private, the Singaporean sovereign wealth fund.
Opinions about AI projects in the business and investment communities are divided. Some believe that the AI bubble will burst, just like the dot-com bubble once did. Others are convinced that this analogy is inaccurate since many AI projects are already generating significant revenue and meeting or even exceeding expectations.
The same article has Amazon’s founder, Jeff Bezos, cautioning AI’s hype and how this makes it so investors have a hard time “distinguishing between the good ideas and the bad ideas.”
In line with his comments, the possibility of betting on an unsuccessful project is quite high. An MIT report notes that only about 5% of companies manage to achieve strong growth. Why is that? Researchers note that most startups fail not because of a lack of technology, but because they never achieve real business adaptation. Companies eagerly develop AI solutions, experiment and show strong enthusiasm for the new tech wave. But in reality, 95% of these solutions never make it into core business processes, according to the report.
On top of that, startups pour large sums into marketing and sales, areas where results are easy to track, while neglecting internal operations, documentation and other essential functions. The outcome is clear: plenty of pilots, but very few real transformations.
I think an illustrative example is CodeParrot, a promising startup from the Y Combinator winter 2023 batch. Their demo to automatically convert Figma designs into production-ready code impressed users, but the startup never managed to achieve stable revenue.
At its peak, the company reached only $1,500 in monthly recurring revenue. After several business model pivots (a journey the founders later described as “being stuck in pivot hell”), they decided to shut the project down.
Another illustrative case is Builder.ai, a British startup that received hundreds of millions of dollars in investments from giants like Microsoft, SoftBank and the Qatar sovereign fund. The company aimed to simplify app development to the level of “ordering a pizza” using AI.
However, again, it failed to build a sustainable business model. The startup spent huge sums on development, marketing and scaling, but did not generate enough consistent revenue from clients and was almost entirely dependent on investor funding. On top of the financial issues, legal problems emerged with the founder under investigation in India. A new CEO tried to save the ship, but the startup ultimately sank.
These are just a couple among the growing list of early-stage AI startups that fail to move from prototype to profitability.
What Have Investors Learned And What Might They Do Differently?
The failures and disappointments of 2025 have made investors far more cautious. Many within the venture market are rethinking the AI hype. Now that there are clear examples of weak business models and unjustified investments, AI startup valuations are shifting toward concrete performance metrics: ARR, profitability and customer retention.
When evaluating startups, it’s better to focus on specific numbers rather than general growth indicators. One safe approach is to follow a bank-like strategy: Provide funding in stages and only if the startup meets concrete goals like lowering costs, turning a profit or proving that growth is sustainable.
Next, it’s important to pay attention to operational processes and model worst-case scenarios. What happens if customer acquisition costs increase or if churn worsens? Does the startup have a plan for these situations? A project’s readiness to handle market challenges should become one of the key criteria for investment decisions.
After a difficult year, investors have realized that a good product and a strong team alone may not save a startup. It’s important to assess the overall market situation. For example, how many companies are actually going public, and how are interest rates and deal valuations behaving? A safer strategy is to watch for triggers and make decisions based on them: Is there activity in the M&A market, and is the liquidity index at a sufficient level? A startup that understands the current market and is prepared for changes earns trust.
The last strategic move that can help avoid losing investments is embedding specialists such as growth engineers, CFOs and customer retention experts into the team. Along with funding, investors should provide expertise, not occasionally, but on an ongoing basis, to prevent operational problems.
The information provided here is not investment, tax or financial advice. You should consult with a licensed professional for advice concerning your specific situation.
Forbes Finance Council is an invitation-only organization for executives in successful accounting, financial planning and wealth management firms. Do I qualify?