On 1 October 2026 Google added a warning against fake author profiles to its people-first content guidance. AI-generated headshots, made-up names and false credentials are now named outright as "a form of deception", according to Search Engine Journal's report of 2 October 2026.
For a UK small business that pays an agency or a content tool for blog posts, this is a quiet change that matters. Plenty of trades, clinics, accountants and letting agents publish articles signed by "Sarah, our property expert", who has a stock smile and never existed. Google has now said in writing that it treats this as deception. This post explains what changed, who is exposed, and a practical way to clean up a small site in an afternoon.
What exactly did Google change in October 2026?
The wording on Google's "Creating helpful, reliable, people-first content" page (accessed 13 October 2026) says Google strongly encourages "adding accurate authorship information, such as bylines", followed by: "However, avoid using deceptive authorship information." It goes on to say that fabricating creator profiles, "such as by using AI-generated headshots, made-up names, or false credentials", to make content look as if human experts wrote it, "is a form of deception". The page adds that "any form of deception makes a page untrustworthy to both users and our automated quality systems".
Search Engine Journal notes that a saved copy of the page from 8 June did not contain this paragraph, and that the same edit added a definition of "main content". Google's page now defines main content as "any part of the web page that directly helps the page achieve its purpose."
Why is Google targeting fake authors now?
Barry Schwartz at Search Engine Roundtable, writing on 6 October 2026, says Google had previously told him it was "not worried about faking author information", and concludes that "times have changed". The practical reason is volume. A content tool can produce a batch of posts, a name and a photorealistic face in an hour, so a byline alone no longer tells a reader anything.
The principle itself is not entirely new. Search Engine Journal points out that Google's Search Quality Rater Guidelines already tell human raters to treat deceptive pages harshly, giving them the Lowest rating. What is new is that the warning now sits in the public guidance aimed at site owners, and it refers to "automated quality systems" rather than human raters.
Does a real byline help you rank on Google?
That distinction matters for budgeting. Do not pay anyone to "add E-E-A-T author boxes" as a ranking tactic. Spend the effort on making sure the names on the site are true and that the content reflects real experience, because that is what the guidance asks for and what customers notice. Search Engine Journal also flags that Google has not explained how its systems detect a fake profile, and no standalone penalty was announced. So the risk is a gradual drag on quality signals, not a switch flipped overnight.
Which UK small businesses are most exposed?
Google's guidance gives extra weight to trust on topics that affect people's money, health and safety, and these are exactly the sectors where UK firms have bought "expert-written" content packages. A typical pattern is a template blog from a marketing platform, signed by a friendly persona with a qualification abbreviation after the name. If you run an accountancy or legal practice, the sector-specific risks are covered in this guide to AI for accountants and solicitors in the UK.
Lower risk: a plumber whose posts are unsigned or signed by the business, or an owner who writes their own updates. Unsigned is not deceptive. Invented is.
How do you tell an honest byline from a deceptive one?
Run each author profile on your site through this checklist:
- If the person is real, works for you and checked the post: keep the byline. Link it to a short bio with their actual role.
- If the person is real but never read the post: have them read and correct it this month, or move the byline to whoever did.
- If a freelancer or agency wrote it and the owner approved it: sign it as the owner or "reviewed by" the owner. That reflects reality.
- If the name is invented: retire the persona. Reassign the posts to a real reviewer or to the business.
- If the headshot is stock or AI-generated: remove it. Use a real photo or no photo.
- If the bio claims a qualification: check it against the relevant register, such as ICAEW, ACCA, the SRA or the GDC. If it cannot be verified, delete the claim.
- If nobody will stand behind a post's claims: unpublish it, or rewrite it with someone who will.
What does a clean-up cost for a small firm?
Take an illustrative Leeds accountancy practice with 60 blog posts, all signed by "Emma Clarke, Tax Specialist", a persona supplied by its content agency. The agency charges £400 a month for four posts. These figures are invented for this example, not sourced, so swap in your own.
- One-off clean-up: a junior spends 3 hours removing the persona and photo and reassigning posts to a "Reviewed by" partner byline. At an internal cost of £25 an hour, that is £75.
- Partner review of the 15 highest-traffic posts: 20 minutes each, 5 hours in total. At a charge-out equivalent of £120 an hour, that is £600 of time.
- Ongoing: 30 minutes of partner review per new post, 2 hours a month, or £240 of time on top of the £400 agency fee.
So the honest version costs the firm roughly £640 a month in fees plus time, against £400 before, plus about £675 once. The extra £240 a month is the real price of having a qualified person stand behind the advice. One new bookkeeping client worth £150 a month covers more than half of it, and two cover all of it. If the posts never generate enquiries, that is a sign the content was not worth publishing at all, fake author or not. The trade-offs of outsourced writing are compared in more depth in this piece on AI content writing services for UK small businesses.
What about AI-written content without a fake author?
Google's guidance describes effort as "the extent to which human work went into creating the content", and says that using generative AI "to produce large amounts of text without manual oversight or curation represents little to no effort". It also says that "attribution or giving credit to other sources doesn't replace the need for original effort", and defines originality as information or perspectives "that aren't already available on other websites."
For a small business, the practical reading is that AI is acceptable as a drafting tool when a real person adds local prices, actual job examples, client questions and a view. That same original material is what AI assistants such as ChatGPT and Gemini tend to quote, which is why honest, specific content also supports AI search visibility.
What are the trade-offs of removing an author completely?
There are three realistic options. A business byline costs nothing and removes the deception risk, but reads as anonymous. A named owner byline builds trust and suits sole traders, but the owner has to read everything. "Written by the team, reviewed by" a named qualified person is the strongest option for regulated advice, and costs the most review time. Choose by stakes: a florist's seasonal tips can carry the shop name, while a mortgage broker's remortgage guide should carry a real adviser.
What should you do this week?
- Monday: visit your blog and list every author name and headshot. Ten minutes on most small sites.
- Tuesday: apply the if/then checklist above. Mark each author keep, reassign or retire.
- Wednesday: remove invented personas and AI or stock headshots. Reassign posts to a real person or the business.
- Thursday: check any claimed qualifications against the relevant professional register and remove anything you cannot verify.
- Friday: email your content agency or freelancer. Require real reviewer names on future posts and ask how drafts are produced.
None of this needs a developer on most WordPress, Squarespace or Wix sites. It is editing author fields and bios.
Sources: Search Engine Journal, 2 October 2026; Google Search Central, people-first content guidance, accessed 13 October 2026; Search Engine Roundtable, 6 October 2026.
Google's Fake Author Warning — FAQ
What did Google add to its guidance about authors in October 2026?
Google added a paragraph to the 'Who' section of its people-first content guidance telling site owners to avoid deceptive authorship information. It names three examples: AI-generated headshots, made-up names and false credentials used to make content look as if human experts wrote it. Google says any form of deception makes a page untrustworthy to users and to its automated quality systems. Search Engine Journal reported the change on 2 October 2026 and dated the updated page to 1 October 2026. The same edit also added a definition of main content, and the paragraph sits alongside Google's existing text on effort and originality. Search Engine Journal notes that no standalone penalty or ranking update was announced with it.
Will adding a real author byline improve my Google rankings?
Not directly. Search Engine Journal points out that Google's Search Liaison said in January 2024 that bylines do not help pages rank better. The October 2026 change is about removing a negative, not adding a positive. A fake profile is now explicitly described as a form of deception that makes a page untrustworthy, so the risk sits with fabrication. A real byline helps customers judge whether to trust the advice, and that trust is what turns a reader into an enquiry. Treat accurate authorship as basic hygiene for both readers and Google, rather than as a ranking trick you can switch on.
Is it deceptive to publish posts written by a freelancer or agency under the owner's name?
It depends on whether the named person genuinely stands behind the content. If the owner briefs the piece, reviews it, corrects it and is happy to defend every claim, putting their name on it is normal business practice and is not what Google's examples describe. The problem cases are invented people, stock or AI-generated faces, and qualifications nobody holds. A safe middle ground is a byline for the expert who reviewed the post, plus a short note on how the content was produced. If you would be uncomfortable explaining the byline to a customer face to face, change it.
Do I have to delete my existing blog posts that have a fake author?
Usually no. Most posts can be kept by reassigning them to a real person who has read and checked them, or to the business itself, and by removing any invented headshot and credentials. Delete only posts that make claims nobody in the business is willing to stand behind, or thin pages that add nothing original. If you change the URL of any post, add a 301 redirect so you do not lose links. Fix the highest-traffic and highest-stakes pages first, such as anything giving financial, legal or health advice, then work through the rest over a few weeks.
Does this guidance mean AI-written content is banned?
No. The new paragraph targets lying about who made the content, not the use of AI tools. The same Google page does say that using generative AI to produce large amounts of text without manual oversight or curation represents little to no effort, and that attribution to other sources does not replace original effort. So AI drafting is fine when a real person adds knowledge, checks accuracy and puts their name to it honestly. What is now explicitly flagged is dressing AI output up as the work of a human expert who does not exist.
