Pangram's Max Spero on why AI detection is harder than 'Real or Fake'
News Source
β’Wed, 02 Sep 2026 17:18:13 +0000
π° What Happened
AI detection startup Pangram has raised $9 million. It also landed a deal with Substack, the writing platform. Substack now shows readers which writers use AI in their newsletters.
CEO Max Spero says telling AI from human is harder than a simple real-or-fake test. He talked about it on TechCrunch's Equity podcast.
The company also released a new tool that detects AI-made images. It wants to be the trust layer for the internet.
π The Backstory
AI tools can now write essays, reviews, and even insurance claims. That flood of fake content is shaking trust online. Job applications and product reviews may come from machines.
A few startups now sell detection as a service. They scan text and images for signs of machine writing.
Detection is a cat-and-mouse game. Each better detector meets a better faker. That is why experts say no tool is 100% sure. The real question may be where 'AI helped' ends and 'AI made' starts.
π― Why It Matters
You read AI text every day without knowing it, in reviews, news, and posts. Tools like these shape how much you can trust what you see. They decide who gets credited and who gets caught.
AI detection startup Pangram has raised $9 million. It also landed a deal with Substack, the writing platform. Substack now shows readers which writers use AI in their newsletters.
CEO Max Spero says telling AI from human is harder than a simple real-or-fake test. He talked about it on TechCrunch's Equity podcast.
The company also released a new tool that detects AI-made images. It wants to be the trust layer for the internet.
AI tools can now write essays, reviews, and even insurance claims. That flood of fake content is shaking trust online. Job applications and product reviews may come from machines.
A few startups now sell detection as a service. They scan text and images for signs of machine writing.
Detection is a cat-and-mouse game. Each better detector meets a better faker. That is why experts say no tool is 100% sure. The real question may be where 'AI helped' ends and 'AI made' starts.
You read AI text every day without knowing it, in reviews, news, and posts. Tools like these shape how much you can trust what you see. They decide who gets credited and who gets caught.