If your agency, your nephew or a chatbot writes your website content, Google’s documentation changed in a way you should know about. In early October 2026, Google updated its guidance on generative AI content to say that it is “critical to manually factcheck and review all AI-generated content” before publishing. Days later it added a warning against deceptive authorship, including fake author profiles. Neither change is a new law, but together they show what Google expects from a business that uses AI to write pages, and what to put in place before the next update lands.
What Google’s pages now say
Search Engine Roundtable reported the first change on 1 October 2026. The guidance page, which we opened on Google’s own site, explains why: generative models predict likely word sequences rather than retrieving facts, so their output can contain inaccuracies, often called hallucinations. Google says AI can help with researching topics and structuring original content. It also says that generating many pages without adding value for users may violate its spam policy on scaled content abuse.
The page lists best practices: focus on accuracy, quality and relevance, and tell readers how content was created, for example by explaining how automation was used. It also reminds site owners that titles, meta descriptions and image alt text can show up in search results, so they need the same review as the body text. For online shops, it adds extra rules: AI-generated product titles and descriptions must be labelled as AI-generated, and AI-generated images need specific IPTC metadata.
On 6 October 2026, Search Engine Roundtable reported a new paragraph in Google’s “Creating helpful content” page. Google says fabricating creator profiles is a form of deception, with examples such as AI-generated headshots, made-up names and false credentials meant to make content seem written by human experts. Google says deception can make a page untrustworthy to users and to its automated quality systems. The article notes that this is a shift from an earlier Google statement that it was not worried about faked author information. Whether any of this ties to the September 2026 spam update is only speculation in the coverage, so do not read it as confirmed.
What this means for a local business website
Most Indian small businesses do not publish thousands of pages. The risk for you is smaller but real: a dentist’s page with an invented statistic, a “doctor” byline with a stock-photo face, or twenty near-identical “Best plumber in [area]” pages produced in an afternoon. These look cheap to a customer first and to Google second.
The fix is not to avoid AI. It is to add a human check, real facts and real people. Use the tool for outlines, rewording and first drafts, then make a named person responsible for what goes live.
A simple review checklist before any page goes live
- Facts: every price, address, timing, qualification, law or statistic is checked against a real document or the owner’s own knowledge. If you cannot source it, delete it.
- Author: the byline is a real person who exists and who agrees with the page. Do not add invented doctors, advocates or “experts”.
- Originality: the page says something specific to your clinic, shop or workshop: your process, your area, your photos. Two pages should never differ only by the city name.
- Metadata: read the title, meta description and image alt text as a customer would. Make sure they are accurate and not stuffed.
- Disclosure: where AI helped produce the page, consider a line on how it was made, in the spirit of Google’s advice to tell readers how content was created.
- Shop products: if you use AI for product text or images on an e-commerce site, follow the labelling and metadata rules on Google’s page.
Keep a short log: who reviewed the page, on what date, what was checked. It takes two minutes and saves an argument later.
AI Integration · FSK Digital Marketing
Know what to automate first, and what it will cost.
From ₹2,999 one-time
Where AI genuinely helps, and where it does not
Used well, AI saves hours on work that does not need judgement: summarising a long call transcript, drafting a reply to a common customer question, turning your price list into a table, or sorting leads. It is much weaker at facts about your business, such as your hours, your doctor’s qualifications or your refund policy. Those should come from you.
That split is how we approach automation at FSK. Our AI automation consulting looks at your day-to-day processes, finds the ones worth automating first and gives you a costed roadmap you can take to any developer. Content is only one candidate. Missed enquiries and manual follow-ups are often a better first target than blog writing, and you can read how we think about automated replies in our post on the ChatGPT business hub and local website chatbots.
The spam-update backdrop
Search Engine Roundtable’s November webmaster report notes that Google finished its September 2026 spam update in October. Search Engine Journal’s timeline puts that rollout at 13 days and 16 hours, from 24 September to 8 October. The documentation edits arrived in the same weeks. You do not need to guess at the connection. The practical point is that Google is spelling out, in writing, what it considers low-effort and deceptive content, so a business that follows the checklist above is on the safe side of those written rules, whatever the next update does.
It is also a reminder that quality is cumulative. One careful page will not rescue a site full of thin ones, and one weak page will not sink a site built on honest, specific content. Review a few pages every week rather than all at once.
Tie this to your ranking checks
If you saw a traffic change recently, do not jump to blaming AI content. The November 2026 core update began on 10 November, and Google advises waiting until it finishes before judging Search Console data. Our guide to measuring local impact after the core update shows how to compare the right weeks. Review your content on its own merits, not as a reaction to a graph.
Frequently asked questions
Is AI-written content banned by Google?
No. Google’s guidance says AI can help with researching and structuring original content. It warns that mass-producing pages with little value may count as scaled content abuse, and it says all AI-generated content must be manually fact-checked and reviewed.
Do I have to tell readers I used AI?
Google’s page recommends telling readers how content was created, such as explaining how automation was used. For product titles and descriptions on shopping sites, it says AI-generated text must be labelled as AI-generated.
Can I use a stock photo as the author picture?
Google’s new paragraph flags AI-generated headshots and invented creator profiles as deception. Use a real author with a real photo and honest credentials, or publish under the business name.
Your next step
Pick your five most important pages, apply the checklist above, and fix anything you cannot back up. If you want help deciding where automation will actually save you time, contact FSK for a conversation about your workflow.
AI Integration · FSK Digital Marketing
Know what to automate first, and what it will cost.
From ₹2,999 one-time
Sources
- Search Engine Roundtable: Google: Manually Factcheck & Review AI-Generated Content
- Search Engine Roundtable: Google Warns Against Deceptive Authorship Information
- Google Search Central: Guidance on using generative AI content
- Search Engine Roundtable: Google Webmaster Report November 2026
- Search Engine Journal: SEO Pulse November 12, 2026

