TL;DR
- Buyers now reject anything that reads machine-written, even one stray line.
- Trust in AI-generated content fell from 73% to 55% in two years (Capgemini).
- The tell is generic copy, and AI makes generic easy to mass-produce.
- Use AI for research and targeting, and keep a human standard on the message.
In May 2026, a single line of copy on a Nike product page set off a public pile-on. Nobody confirmed a machine wrote it. It read like one, and that was enough to make people question whether the brand still had a voice. Fast Company covered the moment and described AI as an “inauthenticity force multiplier.”
That reflex now sits in your prospect’s inbox. Harris Poll research this year found that 73% of people are less likely to trust an ad they suspect was made with AI, and Capgemini tracked a fall in trust in AI-generated content from 73% to 55% over two years.
Your buyers run cold email through the same filter, and a sequence that reads like a template gets skimmed for half a second and deleted.
The common reaction is to blame the AI and write every line by hand again. There is a faster route that keeps your volume and drops the robotic tell. Here is what actually triggers the AI smell in outreach, and how to remove it.
The authenticity backlash reached your inbox
From campaign-level backlash to sentence-level scrutiny
The backlash started with whole campaigns. Coca-Cola drew heavy criticism for AI-generated holiday ads two years running, most recently in November 2025.
By 2026, the scrutiny had moved down to the sentence level. Marketing Week reported that 61% of consumers now believe they can spot AI-generated content, and 42% say their view of it has grown more negative over the past year. People are scanning ordinary copies for the machine to tell, and they punish it when they find it.
The Nike moment, and why a single line did the damage
Nike never ran a deliberate AI copy campaign. One line on a product page was enough. Someone posted a screenshot, it spread, and the conversation turned into whether Nike had lost its voice. The lesson for outbound is direct.
You do not have to automate your whole program to get burned. One templated-sounding sentence can define how a prospect reads the entire email.
Why this matters for outbound teams
Buyers reply to people, and screen out templates
Cold email lives or dies on whether it feels written for the person reading it. McKinsey research found that 71% of consumers expect personalized interactions and 76% get frustrated when they do not get them. B2B buyers hold that bar even higher, because a vague pitch signals you skipped the homework a real deal requires. When your message could have been sent to ten thousand accounts without changing a word, prospects treat it that way.
The gap between what execs think and what buyers feel
There is also a blind spot on the sending side. IAB research covering late 2025 into 2026 found that Gen Z and Millennial buyers feel less positive about AI-generated advertising than ad executives assume, for the second year in a row.
If your team is betting that prospects will not notice or will not care, the data points the other way. The same buyers who mock a machine-made ad on social media are the ones opening your sequence.
What actually makes a sequence sound like a bot
Most “AI-sounding” emails share a small set of tells. Naming them makes them easy to catch in your own copy before a prospect does.
The usual tells
The giveaways are consistent. Overly smooth, symmetrical phrasing that reads like it was polished by a model. Throat-clearing openers such as “I hope this email finds you well.” Value propositions that would fit any inbox without edits.
Personalization that stops at the first name. On that last point, McKinsey is explicit that real personalization has to go past inserting a name or dropping someone into a segment, and operate at the individual level.
A first-name merge tag is not personalization; it is a mail merge.
Why a generic opener convinces no one
An opener like Hi {FirstName}, I saw your company is scaling names with nothing specific. It could go to ten thousand companies untouched, and prospects have seen the pattern thousands of times.
A model can generate it in bulk, which is exactly why it reads as bulk. Specifics are the signal that a human looked at the account, such as a recent hire, a product change, or a number from their last earnings note.
The fix: A five-point human-signal check
Run your best-performing sequence through this list before your next send. If a line fails a point, rewrite it.
- Name one specific thing about the account in the first two lines, such as a trigger event, a role change, or a number they would recognize.
- Cut the throat-clearing opener. Start with the reason you are writing.
- Read the email out loud. If it sounds like a press release, rewrite it in the words you would use on a call.
- Replace the interchangeable value prop with the one outcome that matters to this segment.
- Keep the task small and concrete. Give one clear next step instead of a menu.
Done well, this keeps your volume intact while making the research visible in the words, so relevance and scale stop competing with each other.
Mistakes teams make when they overcorrect
Going fully manual and torching your volume
The panic move is to pull AI out completely and hand-write everything. Pipeline math punishes that fast. You lose the coverage that makes outbound work, and your strongest rep spends the morning researching accounts instead of talking to buyers.
Fake personalization at scale
The other failure does more damage. Bolting a custom first line onto a boilerplate body, where the intro references something trivial and the rest is generic, reads as manipulation the moment a prospect notices the seam. The Journal of Business Research documented an “AI-authorship effect,” where AI-generated attempts at emotional or personal connection can trigger something close to disgust and reduce goodwill toward the brand. Half-personalization lands squarely in that trap.
Let AI do the research and keep your voice
The teams handling this well use AI where it is strongest, on research, account signals, and first drafts, and hold a human standard on the words that reach the inbox. AnyBiz is built for that split. It gathers the account-level signals and personalizes each sequence at scale, so your outreach reads as if a person wrote it while your volume holds.
See how AnyBiz personalizes every sequence at scale. Book your first meeting this week.
