Outbound Links, AI Search, and the New Rules of Authority

A quiet clarification from Google’s John Mueller landed last week that deserves more attention than it got. Asked whether outbound links pass negative signals to linked pages, Mueller confirmed they don’t — unhelpful or low-quality outbound links are simply ignored by Google’s systems rather than treated as a penalty vector. Reassuring, sure. But the more interesting question isn’t what bad links fail to do. It’s what good ones actively build — particularly as generative engines increasingly rely on entity relationships and semantic association graphs to determine who gets cited in an AI-generated answer.

What Mueller’s Answer Actually Reveals About Google’s Link Model

The reassurance that outbound links don’t pass negative signals tells us something structurally important: Google’s link evaluation is asymmetric. Unhelpful links are discarded; relevant, authoritative links are used as signals of topical credibility and entity association. Search Engine Journal’s reporting on Mueller’s response confirms that the concern should never have been about harm — it should always have been about signal value.

For SEO teams, this reframes the outbound linking conversation entirely. Linking out isn’t a liability to manage defensively. It’s an editorial decision that either contributes to your topical authority map or contributes nothing at all. In a generative search environment — where models like Google’s AI Overviews and third-party LLMs build entity graphs from crawled content — being associated with the right sources isn’t just good editorial hygiene. It’s part of how your brand gets encoded into a model’s understanding of a subject.

Practically: audit your outbound links not for risk, but for relevance coherence. If you’re a fintech brand and your blog links to authoritative sources on financial regulation, regional central bank publications, or credible industry bodies, you’re reinforcing your entity’s topical footprint. If you’re linking to generic news aggregators because a writer pulled a quick citation, you’re adding noise.

The Entity Authority Play That Most Teams Are Still Missing

Generative Engine Optimisation isn’t a rebrand of SEO — it’s a recognition that LLMs don’t rank pages, they synthesise entities. When an AI model answers a query about, say, the best logistics software for Southeast Asian SMEs, it draws on its training data’s understanding of which entities are credible, frequently cited, and contextually associated with that domain. Your backlink profile matters less than your citation profile across the open web.

Moz’s recent Whiteboard Friday on travel marketing in 2026, presented by Chloe Osunsami, makes this concrete in a vertical that’s ahead of the curve on this challenge. Travel brands facing AI-driven search displacement — where a user gets a full itinerary from an AI answer without clicking anything — are being advised to invest in digital PR that generates genuine narrative mentions across authoritative publications, not just link-building for ranking signals. Human-first storytelling placed in credible media creates the kind of contextual entity association that generative models draw on.

The same logic applies across verticals. A brand mentioned by name in a well-sourced feature on Nikkei Asia, The Ken, or Tech in Asia carries different entity weight than a brand that appears only in its own blog and press releases. The implication for Southeast Asian brands specifically: regional authoritative media placements — in local-language publications, not just English-language outlets — are underdeveloped as a GEO asset. Bahasa Indonesia coverage in Kompas, Thai-language coverage in The Standard, Vietnamese coverage in VnExpress — these build entity recognition in markets where AI models trained on regional data increasingly shape local search behaviour.

---

Digital PR as Infrastructure, Not a Campaign

The framing that still holds most brands back is treating PR as a periodic campaign rather than a continuous infrastructure investment. If entity authority is accumulated over time through consistent, credible mentions across diverse authoritative sources, then PR cadence directly affects your generative search visibility — not in a sprint, but as a compounding asset.

This is where the tooling conversation matters. The PR stack has matured significantly: AI-assisted pitch generation, real-time media monitoring, and structured distribution channels have made it possible for lean teams to maintain consistent outreach cadences without agency-scale headcount. The operational shift is from “PR push around a product launch” to “always-on entity signal generation.”

For Southeast Asian brands, this has a practical wrinkle: multilingual content PR is still under-resourced relative to its impact. A brand that earns mentions across Thai, Indonesian, and English-language publications simultaneously builds a denser, more geographically coherent entity graph than one that operates in a single language. AI models trained on multilingual corpora — which increasingly power local search in the region — weight this coherence.

The failure mode to avoid: chasing volume over authority. A hundred placements in low-credibility news wires do less for your entity profile than five placements in publications that AI models have learned to associate with reliable information in your domain. Quality thresholds matter more in GEO than they ever did in pure-play link building.

Putting It Together: A Practical Audit Starting Point

If you’re a marketing or growth lead trying to translate this into Monday morning action, three places to start:

First, map your current entity footprint. Search your brand name and core product terms in ChatGPT, Gemini, and Perplexity. Note which competitors appear in AI-generated answers and which publications are being cited. That’s your gap analysis.

Second, review your outbound link architecture with topical coherence in mind — not penalty risk. Are you associating your content with entities that reinforce your authority in your domain? If not, update editorial guidelines for new content immediately.

Third, build a regional media target list that includes authoritative local-language publications, not just the English-language trade press. Prioritise outlets your audience’s AI assistant was likely trained on — established, high-editorial-standards publications with genuine readership, not content farms.

The brands that will dominate generative search results in Southeast Asia in 2027 are mostly building their entity authority right now, quietly, without calling it GEO. The question worth sitting with: is your current PR and content strategy building an entity that AI models will learn to cite — or one they’ll learn to ignore?

---

At grzzly, we work with growth teams across Southeast Asia on exactly this intersection — building search visibility strategies that account for both traditional ranking signals and the entity authority models that generative engines rely on. If your brand is trying to navigate what AI-first search means for your discoverability, we’d enjoy thinking through it with you. Let’s talk