VCs use real-time AI fact-checkers in pitches, sparking groupthink fears
Is this a scandal?
Not yet — an early signal. Noise 31/100, holding steady, across 1 source.
VC firms will likely standardize AI-assisted due diligence workflows because efficiency gains outweigh current prompting flaws, forcing founders to adopt 'model-readable' pitch structures as a baseline requirement.
Noise 31/100 — louder than 97% of tracked AI controversies.
Why it matters
Automated evaluation of startup pitches risks homogenizing innovation as founders optimize for model legibility rather than genuine merit or human judgment.
Key points
- Adam Besvinick reports VCs use Granola and Claude to evaluate pitches in real time.
- Critics argue poorly prompted AI enforces groupthink by prioritizing internet trends over merit.
- Garry Tan defends adversarial AI assistants despite backlash regarding automated evaluation.
- Justine Moore advises technical founders to assume they are pitching to AI proxies.
- Turner Novak claims fundraising viability now depends on pitch legibility to language models.
- Founders are adapting by preemptively rebutting AI flags or demanding in-person meetings.
The story
Venture capitalists are increasingly deploying AI tools like Granola and Claude to transcribe and evaluate founder pitches in real time, according to investor Adam Besvinick. Besvinick reported that poorly prompted models often enforce groupthink by flagging unfounded internet trends during meetings. Prominent investors including Garry Tan and Justine Moore defended the practice, arguing technical founders must assume they are addressing AI proxies. Turner Novak stated that fundraising success now depends on pitch legibility to these models. In response, some founders have begun preemptively rebutting anticipated AI critiques or requesting in-person-only meetings to bypass automated screening. This shift indicates a structural change in capital allocation where algorithmic interpretation influences investment decisions alongside traditional human due diligence.
Who's involved
Warns that real-time AI evaluation promotes groupthink and biases VCs toward unfounded internet trends.
Argues that real-time adversarial AI assistants remain valuable tools despite concerns about implementation quality.
Advises technical founders to assume they are speaking to an AI proxy at least half the time.
States that raising capital is difficult if a pitch is not legible to evaluation models.
How the conversation shifted
Polarity (0–100) from the noise pipeline, sampled over time.
Noise Level
The timeline
Industry figures weigh in on AI legibility
Justine Moore and Turner Novak publicly advised founders to optimize pitches for model comprehension.
Garry Tan defends adversarial AI tools
Quote-replied within one hour asserting value of real-time AI assistants despite Cluely backlash.
Besvinick exposes real-time AI pitch evaluation
Posted thread detailing how VCs use Granola and Claude to fact-check founders mid-meeting, garnering 53.9K views.
The full record
Sources & methodology
- twitter.com — twitter.com
Every claim above traces to these primary items. How we score →
The forecast
VC firms will likely standardize AI-assisted due diligence workflows because efficiency gains outweigh current prompting flaws, forcing founders to adopt 'model-readable' pitch structures as a baseline requirement.
Forecast, not fact — an editorial estimate we score when this resolves.
That's the complete picture as of — nothing more to know right now. We'll update this page the moment it changes.
Follow this story
We keep this page current — no need to check back. We'll send the next real change to your inbox, nothing else.
Tracking this story since September 22, 2026.
Join the Discussion
Discuss this story
Community comments coming in a future update
Be the first to share your perspective. Subscribe to comment.