Pangram's AI-detection Gmail integration and Anthropic's Claude watermark signal a new phase in the AI arms race, with implications for corporate communication and trust.
Alpha Score of 75 reflects strong overall profile with moderate momentum, moderate value, strong quality, strong sentiment.
Pangram, an AI detection service, is building a Gmail integration that scans incoming mail and labels whether it was written by artificial intelligence, CEO Max Spero confirmed to Business Insider. The feature, discovered by researcher Jane Manchun Wong, would also give users an option to move AI-generated emails to spam or trash.
Separately, Anthropic announced it will add a watermark to outputs from its Claude chatbot, making it easier to identify text that the model generated. The two moves signal a growing push to flag AI-written content, even as the technology becomes more common in everyday work.
The shift cuts both ways for corporations. Employees use AI to draft internal memos, summarize meeting notes, and write customer-facing emails. If those messages can be flagged, the risk of eroded trust rises – especially when clients or partners are not told AI was used.
Y Combinator CEO Garry Tan recently addressed the question directly, telling a table of peers that he uses AI for writing and is not ashamed of it. “I sure do, you caught me,” he said. His stance highlights a split in norms: some users are willing to own their AI use, while others prefer it hidden.
Apple, whose Mail app and iMessage platform handle billions of messages, will be part of the conversation as detection tools spread. The company already offers AI writing features through Apple Intelligence, though it does not watermark them. The broader trend could reshape how companies manage internal and external communications.
The arms race between AI generators and detectors is unlikely to end soon. Anthropic has warned that no watermarking system is foolproof. Pangram is still awaiting final permissions from Google. For now, any email sent through Gmail could eventually carry a label its author never expected.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.