Pangram Raises $9 Million to Detect AI-Generated Content as Synthetic Media Floods the Internet

As generative AI floods the internet with text and images, New York-based startup Pangram has raised $9 million to scale its detection technology. The round was led by Menlo Ventures, with participation from Haystack, ScOp, Script Capital, and Cadenza.The funding coincides with the launch of Pangram 4, the company’s latest text detection model, and Pangram Image, its first tool dedicated to identifying AI-generated images. Pangram claims accuracy rates exceeding 99% for detecting AI-assisted writing and mixed human-AI content.Why Detection Matters NowThe volume of AI-generated material online has grown dramatically. From social media posts and marketing copy to academic papers and news-like articles, synthetic content is becoming harder to distinguish from human work.This surge creates real problems:Spread of low-quality or misleading information Challenges for teachers and academic institutions Risks to trust in journalism and publishing Difficulty for platforms trying to maintain content standards Pangram’s goal is to give users and organizations a practical way to identify AI involvement in content.Pangram’s TechnologyThe company focuses on detecting both fully AI-generated text and hybrid content that combines human writing with AI assistance. Its newer models aim for finer-grained analysis, including better performance at the sentence level.With the addition of image detection, Pangram is expanding beyond text to address the full spectrum of synthetic media. This dual approach reflects the reality that AI content now spans multiple formats.Founded by former machine learning researchers from Tesla and Google, the startup has built its reputation on high accuracy claims and targeted use cases in education and enterprise.Use of the New FundingThe $9 million will help Pangram:Improve and scale its detection models Grow its image detection capabilities Expand sales and partnerships, particularly in education and publishing Continue research to stay ahead of rapidly evolving generative AI systems This latest round builds on earlier seed funding and signals strong investor confidence in the need for reliable detection tools.The Competitive LandscapePangram operates in a growing market of AI detectors that includes established players and newer entrants. Demand has been especially strong from universities, schools, and content platforms seeking ways to uphold originality standards.However, the field faces ongoing technical challenges. As language models and image generators improve, the statistical patterns that detectors rely on can become less distinct. Maintaining high accuracy requires constant updates—exactly what Pangram is attempting with its new releases.Limitations and Realities of DetectionNo detection system is perfect. Common issues include:Difficulty with heavily edited or hybrid content Potential false positives that incorrectly flag human writing Varying performance across languages and writing styles The fundamental arms race between generators and detectors Pangram positions its latest models as significant improvements in these areas, but real-world performance will ultimately be judged by users.Bigger PictureThe rise of companies like Pangram highlights a broader shift. As generative AI becomes ubiquitous, tools for transparency and authenticity are emerging as important infrastructure. Some advocate for mandatory AI labeling, while others focus on detection and disclosure technologies.Pangram’s approach centers on giving institutions and individuals the ability to check content rather than trying to stop generative AI itself.