Email in the AI Age: Personalisation Without the Creep Factor

14 January 2026

Two years into the AI era, inboxes tell a clear story: everyone's emails got longer, glossier — and more identical. The tools that promised to make every message personal have mostly made every message sound like the same enthusiastic robot.

Yet email remains the highest-returning channel most businesses have. So here's how the smart operators are actually using AI in 2026.

Where AI genuinely helps

• Segmentation. AI is superb at spotting patterns in your customer data — who buys what, when, and what they ignore. Sending different messages to different groups beats any amount of clever copywriting sent to everyone.
• Timing and testing. Let the machines optimise send times and run subject-line tests. This is maths, and machines are good at maths.
• First drafts, not final drafts. AI gets you from blank page to rough shape in minutes. The last 20% — your stories, your phrasing, your actual opinion — is the part readers subscribe for.

Where it backfires

Fake familiarity ("I noticed you visited our pricing page, John…") reads as surveillance, not service. And fully automated newsletters are now so recognisable that they train readers to skim — or unsubscribe. Personalisation means relevance, not name-insertion.

The test for every email in 2026: would a real person send this to another real person? If not, the delete key knows what to do — and so do your readers.