Entity Optimisation for AI Overviews: Staying Visible in Breaking News

When AI Overviews appear for a news query (which happens in only 15 per cent of breaking news searches) the publication that ranks most often in those summaries is not always the one with the best traditional Google Search ranking. It's the one with the clearest entity definition across the Knowledge Graph, consistent structured data, and verified relationships to topic authorities. This is entity optimisation for AI, and it's reshaping how news organisations think about their digital presence in an era where Generative AI systems like Google Gemini are transforming search results pages (SERPs).

News publishers face a paradox: AI Overviews have a 15% visibility rate for news queries, nearly three times lower than health and science content because large language models can't keep pace with real-time events and hallucinate too frequently in time-sensitive contexts.[1] Yet the small fraction of news that does appear in AI Overviews shows a clear citation preference for publications with strong entity authority. Publishers must invest in entity infrastructure now, not because it solves the traffic crisis today, but because it determines which newsrooms stay relevant as Generative AI systems become more reliable and cite breaking news more often through Google's Search Generative Experience.

(Source: BREAKING! News Thrives in the Age of AI, Shahzad Abbas, March 9, 2026.)

Understanding AI Overviews: Why Breaking News Is Protected

The structural protection of breaking news from AI-generated summaries stems from a fundamental limitation of large language models: they cannot reliably report on events that occurred after their training data cutoff. Large language models show a 20% error rate in AI assistant responses about news topics according to a Reuters study, which is identified as the primary reason AI-generated summaries have not displaced human-reported breaking news in Google Search results.[2]

This creates an unusual dynamic in the search ecosystem. While traditional evergreen content battles for visibility in an increasingly AI-dominated search experience powered by Google Gemini and similar systems, breaking news traffic on Google Discover is up 103% compared to November 2024, even as organic search traffic to publishers has dropped 42% since AI Overviews launched.[3]

The traffic has shifted channels rather than disappeared entirely. Google's algorithm recognises that AI systems cannot be trusted with breaking news, so it routes that demand through Discover and protected search features like the Top Stories carousel. This represents a profound acknowledgement that human journalism remains irreplaceable for time-sensitive information.

Yet when Generative AI systems do cite breaking news, the selection process reveals clear patterns. Journalistic content accounts for 27% of all AI citations overall, rising to 49% for queries that imply any time-sensitivity.[4] This makes breaking news one of the most cited content categories in AI systems despite low AI Overview appearance rates.

The citation window for news content is remarkably narrow but highly valuable. More than half of journalism citations in AI systems come from articles published within the last 12 months, with the highest citation rate for any piece of content occurring within 7 days of publication.[5] This reinforces what news publishers already know: freshness matters. But it also reveals something new: AI systems are actively seeking recent content to cite.

Entity Authority as the Hidden Citation Engine

The mechanics of AI citations operate differently from traditional Google Search rankings. 38% of cited pages in AI Overviews appear in the top 10 Google results for the same query, meaning 62% of AI Overview citations originate from pages ranking outside the top 10.[6] This represents a significant shift from 76% top-10 overlap in July 2025.

(Source: Update: 38% of AI Overview Citations Pull From Top 10 Pages, Louise Linehan, Xibeijia Guan, Ahrefs, March 2, 2026.)

This citation pattern suggests that Generative AI systems are not simply regurgitating the highest-ranking search results. Instead, they appear to be evaluating content based on different criteria altogether: entity clarity, information gain, and semantic relationships rather than traditional Google Search ranking signals.

Consider the case of Forbes, which demonstrates what comprehensive entity authority looks like in practice. Forbes is the only traditional media outlet cited by AI search engines across all 11 major B2B and B2C sectors in analysis, appearing in over 10,000 AI citation mentions in a single study.[7] This cross-sector citation dominance doesn't happen by accident; it results from consistent entity positioning, structured data implementation, and clear topical authority signals.

Traditional SEO strategies are no longer enough to stay visible in an AI-driven search landscape. While ranking for a single keyword used to be the goal, AI engines now fan out user intent into dozens of specific sub-queries. If your content doesn't address these hidden layers of intent, you risk being ignored by the very models your customers are using.[8] The query fan-out process reveals why entity-based content performs better in AI citations. Pages that cover related topic variations through entity linking have higher citation probability than keyword-focused single-query optimisation in SERPs.

For breaking news, this means a story about "UK inflation data" that includes clear entity relationships to the Bank of England, previous inflation reports, and related economic indicators is more likely to be cited than a story optimised solely for the phrase "UK inflation January 2026."

Platform-specific citation patterns add another layer of complexity. ChatGPT cites Reuters, the Financial Times, Forbes, Axios, and Time heavily, while Claude cites the same outlets far less frequently, sometimes 50 times less often than ChatGPT for the same publication.[9] This indicates that entity definition and data availability differs across AI platforms and requires multi-platform optimisation strategy.

Building Entity Infrastructure for Breaking News

Entity optimisation for news publishers extends far beyond traditional SEO tactics. It requires building consistent identity signals across platforms that help Generative AI systems understand who you are, what you cover, and why you should be trusted for specific types of information.

News publishers implementing NewsArticle schema markup, clear byline attribution, and structured data signals are building entity authority that translates to higher citation rates when AI systems do surface breaking news.[10] This infrastructure also improves Discover algorithm performance, creating immediate value alongside future AI citation benefits.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "NewsArticle",
  "@id": "https://www.example-news.com/economy/uk-inflation-january-2026#article",
  "headline": "UK Inflation Falls to 3.2% as Energy Costs Ease",
  "datePublished": "2026-01-15T08:30:00+00:00",
  "dateModified": "2026-01-15T14:12:00+00:00",
  "description": "UK CPI dropped to 3.2% in January 2026, driven by falling energy prices, according to the latest ONS data.",
  "mainEntityOfPage": {
    "@type": "WebPage",
    "@id": "https://www.example-news.com/economy/uk-inflation-january-2026"
  },
  "author": {
    "@type": "Person",
    "@id": "https://www.example-news.com/journalists/jane-smith#person",
    "name": "Jane Smith",
    "jobTitle": "Economics Correspondent",
    "sameAs": [
      "https://twitter.com/janesmith_econ",
      "https://www.linkedin.com/in/janesmithecon"
    ]
  },
  "publisher": {
    "@type": "NewsMediaOrganization",
    "@id": "https://www.example-news.com/#organization",
    "name": "Example News",
    "sameAs": [
      "https://en.wikipedia.org/wiki/Example_News",
      "https://twitter.com/ExampleNews",
      "https://www.wikidata.org/wiki/Q00000000"
    ],
    "logo": {
      "@type": "ImageObject",
      "url": "https://www.example-news.com/assets/logo.png"
    }
  },
  "about": [
    {
      "@type": "Thing",
      "@id": "https://www.wikidata.org/wiki/Q179960",
      "name": "Consumer Price Index",
      "sameAs": "https://en.wikipedia.org/wiki/Consumer_price_index"
    },
    {
      "@type": "Organization",
      "@id": "https://www.wikidata.org/wiki/Q193695",
      "name": "Bank of England",
      "sameAs": "https://en.wikipedia.org/wiki/Bank_of_England"
    },
    {
      "@type": "Organization",
      "@id": "https://www.wikidata.org/wiki/Q1124477",
      "name": "Office for National Statistics",
      "sameAs": "https://en.wikipedia.org/wiki/Office_for_National_Statistics"
    }
  ],
  "mentions": [
    {
      "@type": "Event",
      "name": "ONS CPI Release – January 2026",
      "startDate": "2026-01-15"
    }
  ],
  "isAccessibleForFree": true
}
</script>

Entity disambiguation represents a critical but under-implemented strategy for news organisations. Structured data markup including @id, sameAs, and mainEntityOfPage helps Google (and AI systems) link news content to recognized Knowledge Graph identifiers, with inconsistent entity naming across platforms causing entity fragmentation that reduces AI citation confidence.[11] These features, demonstrated at Google I/O 2024, show how Google Gemini and other systems evaluate content reliability through entity consistency.

A news organisation covering technology should ensure that their entity signals consistently identify them as a technology publication across their website schema, social media profiles, press release distribution, and any third-party platforms. When AI systems evaluate potential sources for a technology news query, consistent entity signals reduce the inference required to understand the publication's topical authority.

The freshness factor creates unique opportunities for news publishers in AI citations. AI-cited content is 25.7% fresher than traditional organic Google results, and pages refreshed within 30 days are dominating AI citations across ChatGPT, Perplexity, and Google's AI Overviews at 3.2x the rate of older pieces.[12] This means publishers that update their breaking news coverage with additional reporting, context, or related entity information can extend their citation window.

(Source: New Study: AI Assistants Prefer to Cite “Fresher” Content (17 Million Citations Analyzed, Ryan Law, Xibeijia Guan, July 28, 2025.)

The Information Gain Advantage

Traditional SEO optimises for ranking. Entity optimisation for AI citations optimises for being the obvious choice when a Generative AI system needs to cite a source. E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) is table-stakes for ranking in traditional search, but in AI search it is even more critical, and information gain (original data that fills information gaps not covered in consensus results) is what gets cited.[13]

For breaking news, information gain often comes from original reporting, exclusive interviews, or first-hand access to events. But it can also come from entity-rich context that connects breaking news to broader patterns, historical precedent, or related developments that other outlets miss.

The press release citation trend illustrates how entity-rich content gains AI visibility. Press release citations by AI search engines increased 5x between July and December 2025 and now account for 1% of all AI citations, the fastest-growing format in the dataset, with PR Newswire, Business Wire, and GlobeNewswire leading citation volume.[14]

Press releases succeed in AI citations because they contain clear entity relationships, structured information, and specific details that Generative AI systems can extract and verify. News publishers can apply similar principles to their breaking news coverage by ensuring that stories include clear entity identification, structured data about key figures and organisations, and semantic relationships to related topics.

YouTube's dominance in AI Overview citations offers another model for entity-based content strategy. YouTube is the most-cited domain in AI Overviews overall and has grown 34% over the past six months, accounting for 5.6% of all AI Overview citations and 18.2% of citations outside the top 100 Google results.[15] YouTube's success stems from rich metadata, clear entity tagging, and comprehensive topic coverage that spans multiple related queries.

Partnership Influence and Direct Integration

The citation landscape includes factors beyond pure optimisation. When AI Overviews do appear for news queries, they favour sources with clear entity authority and consistent citations across platforms, but 29% of news citations in ChatGPT come from publishers that have partnered with OpenAI.[16]

This suggests that direct integration relationships influence citation patterns alongside entity authority. Publishers considering AI partnerships should evaluate them not just for licensing revenue, but for their potential impact on citation visibility across AI platforms.

The broader citation share data reinforces the opportunity for news organisations. News publications make up 14% of all citations in Google's AI products (including AI Overviews, AI Mode, and Gemini) and OpenAI's ChatGPT, with Yahoo, Forbes, and Seeking Alpha as top cited publications.[17] Entity clarity and consistent positioning contribute to citation dominance within this 14% share.

(Source: The Role of News Publications in AI Citations [New Data], March 13, 2026, Vince Nero)

Measuring Entity-Based AI Visibility

Traditional metrics fail to capture entity optimisation success. Click-through rates, organic traffic from Google Search, and keyword rankings matter less than citation frequency, entity recognition accuracy, and cross-platform consistency.

Carly Steven, SEO Leader at Daily Mail, provides perspective on the traffic impact: Daily Mail reports 80-90% clickthrough rate drop when Google AI Overviews appear for news keywords, but this represents a very low single-digit percentage impact on overall traffic because AI Overviews don't typically show for breaking news queries and most organic search traffic is branded rather than informational.[18]

This traffic impact data supports the argument for entity optimisation as a forward-looking strategy rather than an immediate traffic solution. Publishers building entity authority now are positioning for future citation dominance when Generative AI systems like Google Gemini expand into breaking news coverage with greater confidence.

The Sceptic's Argument: Traffic Reality Check

Critics argue that entity optimisation is repackaged SEO with a new name, and that GEO vendors are overselling a tactic that won't move traffic. The evidence partially supports this scepticism: ChatGPT and Perplexity referrals still account for less than 1 per cent of publisher traffic, and licensing deals haven't translated to proportional click-throughs.

The immediate traffic impact remains minimal. When AI Overviews do appear for news queries, the click-through rates drop precipitously, as the Daily Mail data demonstrates.[19] Publishers investing in entity optimisation won't see dramatic traffic improvements in 2026.

But this critique misses the mechanism and the timeline. Entity optimisation isn't about traffic today. It's about citation authority and brand recognition as Generative AI systems become more reliable and expand coverage into previously protected categories like breaking news.

The Discover traffic growth demonstrates that demand for breaking news content hasn't disappeared. It has shifted to different distribution channels. Entity optimisation that improves both Discover performance and future AI citation eligibility represents a hedge against continued search evolution.

Implementation Strategy: Entity-First News Architecture

News publishers should approach entity optimisation as infrastructure investment rather than tactical SEO. This means implementing NewsArticle schema across all breaking news content, creating entity pages for key reporters and news categories, and ensuring consistent entity identification across all publishing platforms. Early testing in Google Labs can provide insights into how the Search Generative Experience will affect content visibility.

Author entity optimisation deserves particular attention for news organisations. Clear byline attribution, author entity pages with consistent biographical information, and structured data connecting reporters to their areas of expertise help Generative AI systems understand the human expertise behind breaking news coverage.

Cross-platform entity consistency requires coordination between editorial, technical, and marketing teams. The same publication entity should be represented identically across the main website, social media profiles, press release distribution, newsletter platforms, and any syndication partnerships.

Content updates and entity linking should become standard practice for breaking news coverage. As stories develop, adding entity-rich context, related background information, and structured data about key figures extends the citation window and improves AI visibility across SERPs.

The Future of News Visibility in AI-Driven Search

The trajectory is clear: Generative AI systems will eventually become reliable enough to cite breaking news at scale. The 15 per cent visibility rate for news queries in AI Overviews will increase as models improve their real-time information processing and reduce error rates in time-sensitive contexts.

When that shift occurs, the publications with clear entity definition, consistent cross-platform presence, and structured data implementation will dominate citations. Building this infrastructure requires significant editorial and technical investment, but the alternative is losing visibility entirely as Google Search continues its evolution towards AI-generated answers through the Search Generative Experience.

The future of news visibility in AI-driven search is not about ranking the same way you did for Google Search. It's about becoming such a clearly defined entity (with consistent, verified information across platforms, structured data that eliminates ambiguity, and demonstrated expertise in specific topic areas) that when Generative AI systems like Google Gemini need to cite a source, yours is the obvious, lowest-inference choice. Publishers that start building this entity infrastructure now will have a structural advantage when AI systems eventually become reliable enough to cite breaking news at scale.