Meta Descriptions, Internal Links, and the AI Fair Use Fight

Google’s search infrastructure is showing its seams in three distinct places simultaneously — and how you respond to each will quietly separate the brands that hold rankings from the ones that wonder why they’re slipping.

Meta Descriptions Are Optional. That Doesn’t Mean Skip Them.

Search Engine Journal reports that Google has officially confirmed meta descriptions carry no direct ranking weight. You could leave every one blank and suffer zero algorithmic penalty. So naturally, some teams will treat this as a license to deprioritise them entirely. That’s the wrong read.

What meta descriptions do control is click-through rate — and in a search landscape where AI Overviews are cannibalising top-of-page impressions, the clicks that make it to your blue link are increasingly precious. A well-crafted meta description is your one shot at a direct pitch to a human who’s already been served an AI summary and is still scrolling. Google will often rewrite descriptions anyway, but when your copy is tight, specific, and matches search intent closely, it tends to survive the rewrite. For Southeast Asian brands running multilingual campaigns — Bahasa, Thai, Vietnamese — this is also one of the few on-page elements where language nuance directly affects whether a regional audience clicks or bounces.

Treat meta descriptions as conversion copy, not SEO hygiene. The distinction matters.

The Silent Equity Drain Your Link Audit Isn’t Catching

Sophie Brannon’s analysis in Search Engine Journal surfaces something most SEO teams don’t catch until it’s already expensive: internal link structures decay passively over time. Not because links break — but because new content, published without deliberate linking strategy, gradually redistributes PageRank in ways that quietly starve your most commercially important pages.

The mechanics are straightforward but easy to miss in practice. Every new blog post, product update, or campaign landing page you publish creates new internal link pathways. If those pages aren’t intentionally pointing equity toward your priority pages, they’re likely diffusing it toward wherever your template defaults — footers, nav items, related content widgets. Over 12 to 18 months of active publishing, a site can effectively bury its own pillar pages under the weight of its own content output.

The fix isn’t complicated, but it requires process discipline. Run a crawl-based equity audit quarterly — tools like Screaming Frog or Ahrefs’ site structure reports will show you which pages are receiving the most internal link equity versus which pages you actually want to rank. Then build an internal linking brief into your content production workflow, not as an afterthought, but as a brief requirement before publication sign-off. For e-commerce brands on Shopee or Lazada’s off-platform blog properties, this is especially acute — category pages need deliberate equity flow, not whatever the CMS decides to auto-link.

---

Google’s AI Fair Use Argument Has Content Strategy Implications

Google published an AI governance paper this week, and the headline position — that training large language models on publicly available web content constitutes fair use — is already drawing criticism from publishers. Search Engine Journal’s Matt G. Southern reports that Google’s paper points to opt-out mechanisms, DMCA takedown processes, and licensing deals for specialised content as sufficient safeguards for rights holders.

Set aside the legal debate for a moment. The strategic implication for anyone producing content is worth examining clearly: if your publicly accessible content is training the models that generate AI Overviews, and those overviews are reducing clicks to your site, you are effectively subsidising the system that’s compressing your traffic. Google’s position — that opt-out controls exist and that paid licensing is available for premium content — suggests the company sees a two-tier content economy emerging: open web content as training data, and paywalled or licensed content as a protected asset class.

For brands in Southeast Asia producing thought leadership, market research, or original data, this is a signal worth acting on now. Content that’s genuinely proprietary — original surveys, first-party customer data, market-specific analysis — has a compounding advantage in an AI-mediated search environment. It’s harder to train on, harder to summarise accurately, and more likely to require a direct citation. Building a content moat around owned data isn’t just a differentiation strategy anymore; it’s a structural response to how AI training economics are evolving.

The Throughline Across All Three

What connects meta descriptions, internal link decay, and AI fair use isn’t obvious at first. But look at the underlying logic: each one is about controlling what happens to your content after it leaves your hands. Who clicks on it. How equity flows through it. Who trains on it. The search environment in 2026 rewards brands that treat these as active governance questions, not passive technical defaults.

The teams winning visibility right now are the ones auditing their internal architecture before it decays, writing meta copy as if every click is earned (because it is), and making deliberate decisions about what content lives behind a gate versus what feeds the open web.

Key Takeaways

• Write meta descriptions as conversion copy targeting post-AI-Overview clicks — specificity and intent-match are what survive Google’s rewrites.
• Audit internal link equity quarterly using crawl data; new content passively starves priority pages if linking strategy isn’t baked into your publishing workflow.
• Original, first-party data and proprietary research is becoming a structural content moat as AI training economics create a two-tier web.

The deeper question worth sitting with: if the open web is increasingly training infrastructure for AI systems, what does a sustainable content publishing strategy actually look like for a brand that needs both discoverability and defensibility? There’s no clean answer yet — but the brands thinking about it now will be better positioned when there is.

---

At grzzly, we work with growth teams across Southeast Asia on exactly this intersection — building search visibility strategies that account for how AI is reshaping what it means to rank, get cited, and drive qualified traffic. If your team is rethinking its content architecture or internal linking strategy, Let’s talk.