In 2022, a typo in a professional email was a mistake. In 2026, it might be the only proof the sender is human.
AI writing polish is now standard-issue. Gmail rewrites drafts. LinkedIn improves posts before publishing. Grammarly goes well beyond typos into restructuring tone and style. The cumulative effect is a strange flattening: every kind of message - from the quick text to the formal proposal - is trending toward the same grammatically correct, well-structured English.
What Grammar Used to Signal
Text style was social data. Lowercase typing among friends signaled informality and closeness. Abbreviations placed you in a community. Typos mid-sentence showed urgency or casualness. These weren't errors - they were deliberate or unconscious signals about the relationship between sender and receiver.
More practically, a message with clean grammar and good structure used to distinguish professional communication from casual conversation. That distinction is collapsing. A law firm's client update and a teenager's Instagram caption can now be grammatically indistinguishable.
Clean, polished writing used to suggest effort or education. Now it suggests the writer clicked a button.
The Authenticity Inversion
Bad grammar is now a stronger signal of authenticity than good grammar.
If you want to communicate "I wrote this myself in 30 seconds," a typo or casual phrasing does that more reliably than a clean sentence. The clean sentence raises a question it didn't used to raise: did a human write this?
This inverts roughly 50 years of professional communication norms. The implicit message of good writing used to be "I care enough to proofread." It now reads, in some contexts, as "I handed this to a machine."
Some communities are already adapting. Group chats and informal online spaces are doubling down on intentionally rough writing - not out of ignorance but as a deliberate signal. If perfect grammar is the AI default, visible imperfection becomes the human marker.
What AI Polish Costs You at Work
For people using AI writing tools daily - a large share of marketers, customer support teams, and content creators - this creates a real tension. AI-assisted output is technically correct and fast to produce. It is also increasingly hard to distinguish from everyone else's AI-assisted output.
This isn't an argument against using these tools. The time savings are real. But treating "polished" as synonymous with "better" is a mistake for anything where personal voice matters: client relationships, team culture, creative work, or any communication where the reader should feel like they're talking to a specific person.
The people who handle this well use AI for a first draft and then edit back toward their own voice - keeping the specific phrase, the slight awkwardness, the sentence rhythm that's actually theirs. That's a distinct skill from either pure writing or pure AI generation, and it's getting harder to fake as the two become harder to tell apart.