Pangram's Max Spero: Why AI Detection Is Harder Than Real or Fake

The Trust Problem with AI Content

The internet has a trust problem, and it’s not just because social media feeds are filling up with AI slop. AI-generated text and images are now making their way into job applications, product reviews, and even insurance claims, leaving platforms and users alike scrambling to figure out what’s real.

A handful of startups have cropped up in the past couple of years to become the “trust layer” the internet needs — including Pangram. The startup recently snapped up $9 million for its AI detection system and landed a partnership with Substack, which is now using Pangram’s tech to show readers which of their favorite authors use AI to write their newsletters. Pangram also recently dropped a new AI image detection tool.

Pangram's Rise and Partnerships

Pangram has rapidly emerged as a key player in the fight against AI-generated content. The startup recently snapped up $9 million for its AI detection system, signaling strong investor confidence in its technology. This funding follows a significant partnership with Substack, which now uses Pangram’s tech to show readers which of their favorite authors use AI to write their newsletters. This move brings AI transparency directly to one of the web’s most popular writing platforms.

Beyond text detection, Pangram also recently released a new AI image detection tool. This expansion broadens its scope to address the growing flood of synthetic visuals online. With backing and key collaborations in place, Pangram is positioning itself as a comprehensive trust layer for the internet, tackling both written and visual AI content.

Max Spero on AI Detection Challenges

For Pangram co-founder and CEO Max Spero, the core issue isn’t simply labeling content as “real or fake.” On the Equity podcast, he explained that the challenge is far more nuanced. The goal isn't just a binary classification, because AI-generated text and images are now making their way into job applications, product reviews, and even insurance claims. This requires a more sophisticated approach than a simple yes-or-no answer.

Spero’s perspective suggests that detection systems must understand context and intent to be truly effective as a “trust layer” for the internet. The complexity lies in distinguishing between different types of AI use and the potential for harm. This is why Pangram’s technology, now partnered with Substack to show readers which authors use AI, is designed to provide a more detailed analysis rather than a blunt instrument. The future depends on these nuanced approaches to navigate the flood of AI content.

Future of AI Detection

Looking ahead, Spero sees the landscape evolving beyond simple binary judgments. He explains that the core challenge is shifting from asking whether something is “real or fake” to understanding the nuanced ways AI is woven into content creation. Pangram’s role, he shares on the Equity podcast, is to build a foundational “trust layer” for the internet. This involves not just flagging fully synthetic text but also identifying partial AI assistance, a growing gray area in everything from job applications to product reviews.

To stay ahead, Pangram is expanding its toolkit. After recently securing $9 million and partnering with Substack to analyze newsletters, the company has also launched a new AI image detection tool. Spero emphasizes that building trust requires a multi-modal approach, ensuring platforms can verify authenticity across text and visual media.