Court filings from the lawsuit between Reddit and Perplexity AI revealed, under oath, that Perplexity's system scores content partly based on how recently it was published or updated. Content that hasn't been touched in roughly three months gets deprioritized when AI systems are deciding what to include in a real-time answer. It doesn't disappear entirely -- but it competes at a disadvantage against content that was updated last week.
This is different from how traditional search works. A well-written page with good links can hold a top Google ranking for years. AI search retrieval is operating on a shorter clock. It's designed to surface current information, and it treats an untouched page as a signal that the source may not be actively maintained. AI systems interpret silence as irrelevance.
What a "freshness signal" actually is
You don't need to rewrite your clinical pages. The medicine hasn't changed. What you need is a signal that a human being has looked at this page recently and confirms it's still current. AI search can read that signal. It can't evaluate clinical accuracy.
The most effective freshness signal for a NaPro or FABM practice page is a "last reviewed" line with your name and credentials: Last reviewed March 2026 by Dr. [Name], NaProTechnology Medical Consultant. This does two things at once. It tells AI systems the page is actively maintained. And it adds an author attribution that Google's current algorithm weighs heavily for health content -- the kind of attributed, credentialed authorship that AI systems look for when deciding whether a source is trustworthy enough to cite.
Three other lightweight ways to update a page
Add a new patient question to your FAQ section. The questions patients ask you in 2026 are different from the questions they asked in 2022. Adding two or three new ones to an existing page updates it without changing anything that's already there.
Add a citation to recent published research. The RRM field has produced peer-reviewed outcomes data in the last few years -- studies on PCOS, endometriosis, recurrent pregnancy loss, published in journals like Frontiers in Medicine and JRRM. If that research supports what your page already says, adding a citation to a 2024 or 2025 study both resets the freshness clock and strengthens your page's credibility as a clinical source.
Add a date to a statistic. "Studies show NaProTechnology achieves comparable pregnancy rates to IVF for couples with unexplained infertility (Boyle et al., 2025)" is more valuable to an AI search system than the same sentence without attribution. The date tells the system the information is current. The citation tells it there's a verifiable source.
None of these require rebuilding anything or revising your clinical voice. They require the same periodic review discipline that a medical journal applies to its published content. That discipline is exactly what AI systems are now built to reward.
Frequently asked questions
Why does AI search ignore older content?
AI retrieval systems favor recently updated content because freshness correlates with accuracy in fast-moving fields. Pages with no modification signal after 90 days are deprioritized in favor of content that shows a pattern of active maintenance, even if the underlying clinical information has not changed.
What is the easiest way to update a freshness signal without rewriting content?
Adding a visible 'Last reviewed: [month year]' line near the top of the page and updating the HTML lastmod timestamp is the minimum effective change. Adding one new FAQ question or citing a recent study from the past 12 months also resets the signal with minimal editorial effort.
Does adding a 'last reviewed' date actually affect search rankings?
Both Google and AI retrieval systems read structured date signals including schema.org dateModified and visible on-page dates. A credible review date -- especially when accompanied by a small content change -- tells crawlers the page is actively maintained and clinically current.
How often should NaPro practice pages be refreshed for AI search?
A 60-90 day refresh cycle is sufficient for most clinical content pages. The goal is not constant rewriting but demonstrating ongoing stewardship. Quarterly FAQ additions, new citation links, or updated statistics satisfy the freshness requirement for most AI retrieval windows.
Can AI search visibility drop even if my Google rankings stay stable?
Yes. AI retrieval applies freshness filters on top of Google ranking signals. A page can hold its Google position while being excluded from AI-generated answers if it lacks a recent modification signal, because the two systems apply different scoring criteria.