Tell-Tale Signs of AI Use in Y Combinator Applications
Clear indicators reveal founders are using AI to craft their startup applications.
1AI's Impact on Y Combinator Applications
Y Combinator, the prestigious accelerator behind companies like Reddit and Airbnb, has noticed increasing signs of AI-generated applications. Since 2005, YC has invested in over 5,000 startups, requiring founders to explain their idea selection, competition analysis, and revenue models. However, application length has grown by 60% in three years, strongly suggesting AI influence. Partners observe unnatural linguistic patterns becoming prevalent in submissions.
2Linguistic Shifts in Applications
YC partner Tyler Bosmeny noted applications are becoming significantly longer. Pete Koomen shared data showing usage of the word 'wedge' jumped from under 1% in spring 2025 to over 20% by summer 2026. Em dash usage also surged, appearing in more than 50% of applications by summer 2025. These changes coincided with the release of Anthropic's Opus 4 model, suggesting AI's growing role in application writing.
3AI's Signature Vocabulary
YC co-founder Paul Graham highlighted in 2024 how the word 'delve' became a telltale sign of ChatGPT use. Researcher Philip Shapira's data showed 'delve' appearances in publications skyrocketed from near zero to 18,000 instances between 1990-2024. Graham emphasized such words aren't used in spoken English but appear when people 'want to sound clever' in writing. This pattern helps identify AI-generated content.
4Radical Style Shifts in Submissions
In May 2024, Graham observed founders adopting an unnatural 'hard-hitting journalistic style' in emails, which he attributed to AI generation. Koomen's data showed 'delve' usage peaked at just 1.25% in winter 2024 before Graham's viral tweet about it. After the exposure, founders apparently switched to different AI markers, demonstrating how quickly these patterns evolve as detection methods improve.
5Implications for the Application Process
YC leaders believe AI-generated content compromises the authenticity of applications. Graham expressed feeling 'lied to' when reading AI-crafted submissions. The trend toward longer, more polished applications makes evaluation challenging. Consistent patterns in word choice and punctuation raise questions about authenticity, potentially requiring new evaluation criteria in startup ecosystems.
6Historical Context and Comparison
Silicon Valley application processes have continually evolved. The early 2000s favored concise applications, while the 2010s saw more detailed narratives. With AI proliferation, 2020s applications break records in both length and complexity. However, this often showcases AI's capabilities rather than founders' genuine abilities. The shift reflects broader technological impacts on entrepreneurship.
7Potential Industry Consequences
Rising AI-assisted applications create two major challenges for startup ecosystems. First, distinguishing genuine founders from AI-generated applications becomes difficult. Second, investors' ability to identify truly promising teams diminishes. How institutions like Y Combinator adapt will significantly impact the future of startup culture and investment strategies.
8Key Takeaways for Startup World
AI's influence on applications demonstrates technology's transformative power in entrepreneurship. Y Combinator's data reveals both language models' impressive mimicry capabilities and the ethical questions they raise. The authenticity and creativity central to startup culture face new challenges in the AI era. This situation demands fresh strategies from both accelerators and founders to maintain meaningful evaluation processes.
Original article: Entrepreneur – Latest — Y Combinator Başvurularında AI Kullanımının İşaretleri