Observed Signal · Jun 20, 2026 · Product Test · Source: t3n · Impact: 3/5 · Sentiment: Neutral
Pangram Tool Tested for Detecting AI‑Generated Texts
Major German newsrooms used the AI-detection tool Pangram to check whether politicians’ speeches and guest articles were written by humans or generated by AI. Pangram is a product of startup Pangram Labs, founded in 2024 by Stanford alumni Max Spero and Bradley Emi. The company trains its detector using a technique called “negative mining,” where cases the model misclassifies (false positives and false negatives) are added to the training set to improve accuracy. t3n evaluated Pangram on its Tool Time show and attempted to trick the detector with a deliberately AI-generated speech. Media reports cited include Frag den Staat, Die Zeit and Focus, the latter reporting a claim that ~9% of Bundestag speeches in 2026 were fully AI-generated.
Tool-level developments in AI‑text detection matter to publishers, journalists and platforms verifying content authenticity; Pangram’s use by major newsrooms and its training method (negative mining) affect media fact-checking and public trust in political communication.
Track DIE ZEIT Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Pangram is a detection tool produced by startup Pangram Labs.
- Pangram Labs was founded in 2024 by Max Spero and Bradley Emi (both Stanford graduates).
- Pangram’s detector is trained using a 'negative mining' approach that enriches training data with misclassified examples (false positives and false negatives).
- German outlets (t3n, Die Zeit, Frag den Staat, Focus) used Pangram to investigate whether politicians’ texts were AI‑generated.
- Focus reported research claiming roughly nine percent of Bundestag speeches in 2026 were fully AI‑generated.
Connected Companies & Entities
3 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
How to Recognize AI-Generated Texts
t3n summarizes a MeisterPrompter podcast episode that outlines clues and limits for spotting AI-generated German texts. Hosts Susanne Renate Schneider and Stella‑Sophie Wojtczak note recurring signals such as overused marketing phrases, prominent "if-then" sentence structures, and the use of longer em-dashes in bullet lists (attributed to English‑heavy training data). The piece warns that no single indicator is definitive and that automated AI‑detectors frequently misclassify content. It suggests practical steps for writers—like ban lists to avoid AI buzzwords—and points listeners to the podcast for more examples. The article also discloses it was produced with t3n’s internal AI tool.
Pangram launches Chrome extension to label AI content
Pangram, a startup, launched a Chrome extension that labels AI-generated or AI-assisted text in real time across social and publishing feeds (X, LinkedIn, Substack, Reddit, Medium). The extension provides inline labels, a sidebar tracking the share of AI content in a user’s feed, and a right-click scan for selected text. Pangram says its token-level classifier — trained on millions of paired human/AI documents and calibrated to strongly penalize false positives — yields a claimed 1-in-10,000 false positive rate and enforces a 50-word minimum for reliable detection. The company collaborated with the Internet Archive on prevalence estimates and with NewsGuard on identifying “AI Content Farm” sites. Pangram’s commercial bet is a $20/month consumer subscription, with expansion plans (image/video detection and potential API use) discussed as possible growth paths.
Pangram's Max Spero Discusses AI Detection Challenges
TechCrunch's Equity podcast features Pangram co-founder and CEO Max Spero discussing the difficulties of detecting AI-generated text and images. The interview highlights how AI-generated content is entering job applications, product reviews, and insurance claims, creating a broader internet trust problem. Pangram recently raised $9 million for its AI detection system and partnered with Substack, which now uses Pangram's technology to show readers which newsletter authors use AI. The startup also released a new AI image detection tool. Spero explores the promise of AI detection tools and the line between AI-assisted and AI-generated content.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.