How to Spot AI-Generated Emails in Your Inbox

As AI-generated content becomes increasingly prevalent in everyday communications, NBC News chief tech analyst Joanna Stern has published a practical guide to identifying whether an email was written by a bot or a human. The article, published Aug. 6, offers readers concrete strategies for detecting “AI slop” in their inboxes.

The ‘AI Slop’ Phenomenon

“AI slop,” also known as slop content, refers to digital content made with generative artificial intelligence that is perceived as lacking in effort, quality, or meaning, and produced in high volume as clickbait. The term has a pejorative connotation similar to spam, and “slop” was selected as the 2025 Word of the Year by both Merriam-Webster and the American Dialect Society, according to Wikipedia.

Stern, who joined NBC News in March as part of a new partnership model between the news organization and independent journalists, opens her analysis with a personal anecdote: “I wasn’t really annoyed by AI slop until a few weeks ago. I received feedback from a company I was working with, read the first few lines and immediately knew.” Her article, subtitled “Getting ‘AI slop vibes’? Here are the best ways to figure out if that email was written by a human,” is part of her broader coverage of AI’s impact on daily life.

How to Detect AI-Generated Emails

Experts have identified several telltale signs that an email was written by AI rather than a human. According to WarmySender, a cold email deliverability platform, the key signals include:

Uniform sentence length: AI defaults to 15-20 word sentences with consistent complexity, while human writing naturally varies between short and long sentences.
Predictable phrasing patterns: AI models overwhelmingly start cold emails with “I noticed…” openers, creating a recognizable fingerprint.
Absence of contractions: AI often writes “do not” instead of “don’t,” producing text that feels overly formal.
Symmetrical paragraph structure: AI produces evenly-sized paragraphs, while human writing is messier and more varied.
Formal tone without personality: AI defaults to professional but bland language, lacking the small imperfections that signal genuine human thought.
Hedge-free assertions: AI makes confident claims without hedging, while humans naturally use qualifiers like “I think” or “in my experience.”

AI Detection Tools and Their Limitations

For those who want more than intuition, several AI detection tools have emerged on the market. Services like GPTZero, QuillBot’s AI Content Detector, and Grammarly’s AI Detector claim to identify AI-generated text from models such as ChatGPT, GPT-5, and Gemini. However, these tools have known limitations, including false positives and varying accuracy rates. No AI detector is 100% accurate, and results should not be used as a final verdict. Human judgment remains essential.

How Email Providers Detect AI Content

Beyond consumer tools, Gmail and Outlook have developed sophisticated machine learning systems to identify AI-generated content patterns. According to Winnr, these systems analyze statistical text properties like perplexity scores and burstiness, structural patterns such as template fingerprints and the “I noticed” opener, and behavioral signals like volume velocity and personalization uniformity.

Email providers don’t directly detect “AI-written” content, but rather the patterns that AI-generated email creates: identical sentence structures across thousands of messages, unnaturally perfect grammar, and suspicious volume spikes.

The Trust Problem

The rise of AI-generated emails has created what researchers call an “authenticity gap” in digital communication. Research shows only about 40% of recipients perceive AI-written messages as sincere, compared to significantly higher rates for human-written correspondence, as noted by Metaintro. This erosion of trust has implications for workplace communication, where colleagues increasingly sense when messages are templated or automated.

The Canadian Bankers Association warns that AI-generated scams are also on the rise, with scammers using AI to create convincing phishing emails, voice-cloned calls, and fake video messages. The organization advises consumers to watch for unnatural or overly formal language, be suspicious of unsolicited requests for information, and verify requests through official channels.

What’s Next

As AI models become more sophisticated at mimicking human writing, the line between human and machine-generated communication will continue to blur. Stern’s guide represents a growing consumer awareness of AI’s presence in everyday communications, and the practical skills needed to navigate an increasingly automated world. The question moving forward is not just how to detect AI slop, but how to preserve authentic human connection in an age of machine-generated content.