Want to appear in AI search results? Target top media publications.
AI large language models (LLMs) like ChatGPT are rapidly becoming a default research assistant for journalists, analysts, tech buyers and even your mate Steve who still thinks blockchain is a soft drink.
And guess what? One of the key sources that AI LLMs learn from is media publications.
What this means for brand discovery and media strategy
Google search links have dropped 30% over the last year as more people turn to AI chatbots like OpenAI’s ChatGPT and Google’s Gemini for answers. And honestly, why wouldn’t they? Letting ChatGPT handle your question instead of wading through a wall of links is a no-brainer for most.
For marketing and PR professionals this changes the traditional world of organic brand discovery from a world of keywords and backlinks to the murky realm of AI content generation.
Sources like Wikipedia, social media, G2, academia and media publications directly feature in the results surfaced by Chatbots like ChatGPT. Recent research from answer engine platform Profound revealed that close to half of ChatGPT’s responses draw from Wikipedia, while Reddit contributes just over 11%, and traditional media outlets make up around 7% of responses. But guess what? Media sources make up a substantial component of Wikipedia and Reddit references too.
So, whether it’s directly or indirectly, media sources play an important part in ChatGPT’s ecosystem. This is because chatbots rely on trusted sources to give users the correct answers, and in a world filled with misinformation, media publications offer AI models the chance to deliver accuracy. If you’re in tech marketing, PR or comms, understanding which media outlets are being read, scraped, licensed, paraphrased (or even stolen) by AI matters a lot.
But the question I’ve been asking myself is: which technology publications feature the most in ChatGPT responses? So, I turned to ‘the best’ source for the answer. ChatGPT.
It’s not just about media sources
Before I get into the list, it’s only fair to put the value of media mentions into context. (I work for a PR and marketing agency, so me banging on about PR mattering is hardly surprising.)
Even if your brand features heavily in media coverage, that media citation getting used as a source by AI doesn’t mean that your brand will get name-checked in AI results. That’s because LLMs like ChatGPT are ingesting, indexing and sometimes summarising this content in real-time. But if your brand appears in these outlets, your messaging has a higher chance of showing up when someone asks ChatGPT: “what’s the top AI trend this year?”
It’s also valuable to note the other crucial factors that influence AI visibility, such as:
Schema markup = robot whispering: Adding structured data (e.g. FAQ, article, product schema) to your website pages gives models and search engines clear cues on what your content is about.
FAQs = free prompts: LLMs are trained to answer questions. If your content already does that, clearly and concisely, you’re ahead. Adding relevant FAQs to landing pages and media coverage which feature SEO keywords is a win.
Quality beats keyword stuffing: As with SEO, LLMs conducting AI searches look for trustworthy, well-structured content. That means credible sources, skimmable layout and clear explanations. Evidence-based, well-written thought leadership is still a winner.
Methodology: how we extracted the top media outlets
Undoubtedly though getting in the right media publications matters for your chances of appearing in AI answers. So back to my central question: which tech publications are more likely to appear in AI search results?
In all honesty, I can’t tell you with total accuracy. (Sorry!) Unfortunately, OpenAI and ChatGPT don’t publish their full list of citations they use, so there isn’t a neat CSV of every URL they scrape.
I can however use data from OpenAI/ ChatGPT’s Deep Research tool and desk research to make a pretty good guess. I used this data to infer what they are using. This research focused on:
- Techmeme data: Media outlets that appear frequently in tech headlines and are cited by other media like Wiki sites (measured as “presence” and “leadership”).
- LLM content deals: who Open AI have done deals with. We know who’s signed licensing deals with OpenAI (and others), so their content is likely prioritised, including Newscorp (Wall Street Journal, The Times and Financial Times), Conde Nast (Wired, The New Yorker), Associated Press, Vox Media (The Verge) and more. Undoubtedly, paywalls and restricted access by publications affect the sources AIs will cite.
- SEO and domain authority: High-ranking sites tend to show up more in AI answers – models mimic top search results. Many of those results are media publications who tend to have stronger SEO authority/trust scores
- Citation patterns: Checking out the content surfaced in AI tools like Perplexity, Claude, Poe and ChatGPT.
- Qualitative analysis: Looking at how often these sources appeared in public dev communities like Hacker News, Reddit and Substack posts.
The results highlighted around 350 publications which were narrowed down to create the final 100. It’s not a perfect science but it’s a pretty solid map.
The Top 100 tech media sources referenced by AI models
According to my analysis, these are the tech publications most likely shaping AI-generated responses, ranked by citation frequency, prominence in AI-friendly ecosystems and SEO authority.
Top 50 (Heavyweights)
- Bloomberg
- TechCrunch
- Wired
- The Verge
- The Wall Street Journal
- The New York Times
- Reuters
- Financial Times
- 404 Media
- The Information
- CNBC
- The Washington Post
- Ars Technica
- BleepingComputer
- The Guardian (Technology)
- 9to5Mac
- 9to5Google
- NPR (Tech)
- BBC (Tech)
- NBC News (Tech)
- Axios (Tech)
- Android Authority
- Krebs on Security
- The Atlantic
- Platformer
- Daring Fireball
- Politico (Tech)
- Semafor (Technology)
- MIT Technology Review
- Variety (Tech)
- CBS News (Tech)
- Windows Central
- Techdirt
- Micah Flee’s Blog
- MacRumors
- Engadget
- Citation Needed
- Simon Willison’s Weblog
- Pew Research Center (Tech)
- Ming-Chi Kuo (Medium)
- The Record
- Fortune (Tech)
- IGN (Tech)
- The Free Press
- Disruptionist
- ProPublica (Tech)
- Forbes (Tech)
- Rolling Stone (Tech)
- Associated Press (AP)
- CoinDesk
51–100 (Niche Influencers)
51. The Block
52. Rest of World
53. Nikkei Asia
54. Business Insider (Tech)
55. SiliconANGLE
56. Tom’s Hardware
57. South China Morning Post (Tech)
58. Cointelegraph
59. GeekWire
60. CNN (Tech)
61. The Hollywood Reporter (Tech)
62. SemiAnalysis
63. The Economic Times (Tech)
64. CNET
65. ZDNet
66. Gizmodo
67. TechRadar
68. PCMag
69. Mashable (Tech)
70. The Next Web (TNW)
71. VentureBeat
72. XDA Developers
73. Android Police
74. AppleInsider
75. Macworld
76. Dark Reading
77. CSO Online
78. SC Media
79. InfoWorld
80. Computerworld
81. eWEEK
82. AnandTech
83. Futurism
84. CIO.com
85. Tech in Asia
86. IEEE Spectrum
87. Digital Trends
88. Wccftech
89. PCWorld
90. Network World
91. TechSpot
92. BGR (Boy Genius Report)
93. The Intercept
94. InformationWeek
95. Quartz
96. Fast Company (Tech)
97. Electrek
98. EE Times
99. Silicon Republic
100. Tech.eu
Sector/subdomain leaders: Who rules what?
- AI & Machine Learning News: Wired, MIT Technology Review, VentureBeat, Axios AI+, IEEE Spectrum
- Cybersecurity & InfoSec: Krebs on Security, BleepingComputer, Dark Reading, The Record, Ars Technica
- Enterprise/Cloud Computing: SiliconANGLE, InfoWorld, Computerworld, ZDNet, eWEEK, CIO.com
- Consumer Tech: The Verge, TechRadar, CNET, Tom’s Guide, Engadget, Gizmodo, PCMag
Where this data comes up short (but still helps!)
It’s worth flagging a few caveats. Firstly, we don’t get direct citations from ChatGPT or Gemini, and since AI models are known to hallucinate, there’s always a bit of uncertainty in its output. Secondly, most models are trained on frozen datasets. For example, ChatGPT-4 web cut-off was late 2023 unless the content is licensed, so there’s a lag in what they know – unless the information is publicly available on the web.
There’s also an inherent English-language and US-centric bias in the data, though we’ve done our best to include global sources where possible. And finally, once you move beyond the top 40 media outlets, influence becomes harder to measure reliably, especially when you start breaking down influence by sector.
Why this data still works:
This list is more of a directional heat map to help prioritise outreach for media coverage. It can also inform your SEO, FAQ and schema strategy to boost your changes of showing up in AI-generated answers. Plus, it helps validate where you should be targeting, publishing (and repurposing) content for your earned PR strategy. And if you’re exploring paid media placements, it’s a useful guide for that too.
Comms + marketing takeaways: how to use this list
So, with everything laid out, let’s talk about how you can best use this information:
- Pitch smarter: Prioritise outlets that feed the machines. That doesn’t mean other publications aren’t valuable but focus will increase chances of AI search mentions.
- Audit AI visibility: Ask ChatGPT/Claude etc questions your customers might. See what comes back. Figure out the key questions prospects and other influencers are likely to ask and create content around them for FAQs to increase the chances of discovery. While we don’t currently know what questions people are typing into ChatGPT, tools like SEMRush and KWFinder provide a good indication of what is likely to be asked based on current search data.
- Build better landing pages: Add FAQs, schema and actual answers to actual questions on your website and into media content. Consider adding a few FAQ responses at the bottom of key landing pages and look to create FAQs pages which answer a bunch of questions in one place.
- Think beyond keywords: Write like you’re explaining something to a smart human who hasn’t had their coffee yet. AIso, focus on quality information instead of a selection of keywords. If the AI sees a valuable response to a question that’s coming from a trusted media source, then you are halfway there.
- Invest in thought leadership: High-authority content = backlinks = media coverage = AI discoverability. The world of SEO has been moving in this direction too. Think about the value your content offers the reader, not your sales pitch. Quality beats quantity.
FAQs
Q: Does ChatGPT tell you which sources it uses? A: No. It doesn’t cite unless prompted. Even then, it might hallucinate citations. That’s why we infer based on available data.
Q: How can I boost my brand’s visibility in AI-generated content? A: Get featured in Wikipedia and high-authority outlets (especially those with AI licensing deals), add schema markup to your site, and create well-structured FAQ content.
Q: Does traditional PR still matter in an AI world? A: Absolutely. Media mentions in the right outlets have downstream effects across SEO, AI content, analyst briefings, and more.
Q: Is this list final? A: No. It’s a 2025 snapshot. There isn’t a definitive list. New outlets rise; others fade. But the patterns (licensing, SEO, content structure) will hold.
Final thoughts: Robots don’t sleep, but they do read
Your media strategy is more than just what people read. It’s now about understanding the sources the machines learn from. If you want your story to be repeated, paraphrased and echoed in a thousand AI-generated summaries, you need to get in front of the sources they trust most.
Because let’s face it: if AI states it, people believe it. So, make sure what it says comes from you.
Want to find about more? Contact us today at hello@rlyl.com
