Elon Musk said on 6 November that “agentic internet traffic already exceeds human traffic”, attributing the figure to Cloudflare, in an exchange at the Baron Investment Conference that the clip’s poster says included Ron Baron. Cloudflare Radar showed bots at 57.4% of web-content requests in June. A clip of the remarks was posted on X (the clip), and Musk reposted it with a comment of his own (his post).
Cloudflare’s own data broadly back the first half of that. In June, Cloudflare Radar showed automated traffic at 57.4% of web requests against 42.6% from people, and chief executive Matthew Prince wrote that bots had “now passed human traffic online for the first time in the Internet’s history”.
Business Pill · AI Agents Explained
A short explainer on AI agents, software that carries out multi-step tasks. It teaches the general idea only and does not discuss this story.
The key insight:Musk said agentic internet traffic already exceeds human traffic, citing Cloudflare and hedging the attribution; Cloudflare Radar showed bots at 57.4% of web-content requests in June, and Prince has forecast automated traffic at 1,000 times human within five years.
According to our transcript of the clip, Musk said: “my understanding is that actually agentic internet traffic already exceeds human traffic”. He went on: “If I read that correctly from Cloudflare, and I believe Cloudflare predicted that in five years, agentic AI traffic will be a thousand times more than human traffic. And if anything, I think that is underestimating the amount of agentic AI traffic.” He hedged the attribution himself (“if I read that correctly”).
In the same exchange, discussing Starlink, Musk said: “you’re roughly doubling revenues per year at large scale”, adding that this “used to be a big deal” and that “with AI, it does exceed that”. His repost of the clip read: “Roughly doubling revenue at scale for a large company used to be something special, but super intelligence makes that look slow!” The doubling remark referred to Starlink, not Cloudflare. Our coverage of his other Baron remarks is here and here.
Cloudflare Radar’s split of requests for web content in June 2026, as reported: 57.4% automated, 42.6% human. Bots ran about 1.35 times humans; 1,000x in five years would need about 3.75x growth a year (our arithmetic).
What Was Posted on X
Roughly doubling revenue at scale for a large company used to be something special, but super intelligence makes that look slow!
Elon Musk’s post at 21:01 UTC on 6 November, quoting a clip of his remarks at the Baron Investment Conference: “Roughly doubling revenue at scale for a large company used to be something special, but super intelligence makes that look slow!” In the clip, the doubling remark refers to Starlink’s revenue.
What Cloudflare Has Said
On 3 June, Prince posted on X (his post): “Welp, that happened faster than I predicted. Thought it would be end of 2027, then early 2027, but agentic traffic growing so fast that bots have now passed human traffic online for the first time in the Internet’s history.” He linked Cloudflare Radar’s bot-versus-human chart (Radar). In a follow-up reply he called the data “a bit messy”.
The Radar figures at the time showed automated systems at 57.4% of requests for web content and people at 42.6%, as reported by SiliconANGLE (SiliconANGLE) and PPC Land. The thousand-fold forecast came later, on The Verge’s Decoder podcast, published on 26 September (The Verge). Prince said there, as quoted by The Next Web: “Five years from now, we think that automated traffic will be 1,000 times human traffic online”. PPC Land reported that he attached a caveat to the extrapolation.
Our Read of the Numbers
By our arithmetic, a 57.4% to 42.6% split means automated requests ran about 1.35 times human requests in June. Reaching 1,000 times human traffic in five years would require that ratio to multiply about 742-fold, or roughly 3.75 times a year, if human traffic stays flat, as Prince has said he expects. Musk said he thinks the 1,000x figure is an underestimate; he gave no number of his own.
Two definitions differ. Prince’s June post counted bots, which include crawlers that collect data for AI training and search as well as agents acting for users; Musk spoke of agentic traffic. The Radar split also measures requests for web content on Cloudflare’s network, which Prince has put in front of more than 20% of the web, not all internet traffic. Agents acting inside company systems raise their own security questions (our Varonis piece).
The Structural Read
As we read it, Musk’s first claim matches what Cloudflare’s own chief executive posted in June; his addition is the view that the 1,000x forecast is too low.
As we read the definitions, Cloudflare’s June split counts all bots, including training crawlers, while Musk spoke of agentic traffic; the two are not the same measure.
As we read the numbers, a 57.4% to 42.6% split is about 1.35 to 1, by our arithmetic, so the 1,000x forecast implies roughly 3.75x growth a year in that ratio if human traffic stays flat.
Matthew Prince, Cloudflare chief executive, on X, 3 June 2026
“Welp, that happened faster than I predicted.”
Three Implications
FOR CLOUDFLARE’S MEASURE Radar counted bots at 57.4% of requests for web content in June, as reported by SiliconANGLE and PPC Land.
FOR THE FORECAST Prince said, as quoted by The Next Web from The Verge, that automated traffic will be 1,000 times human traffic online five years from now.
FOR MUSK’S VIEW Musk said he thinks the 1,000x figure is an underestimate; he gave no figure of his own.
The analysis says: “The surface, not the model, governs autonomy.”
As we read it, the web pages and APIs that bots and agents request are such surfaces, and Cloudflare sits in front of many of them. The analysis does not discuss Cloudflare, Musk or this traffic data.
What Is Not Established
We did not read a current Cloudflare Radar figure for November; the 57.4% share is June’s, as reported at the time. We did not find a Cloudflare breakdown that isolates agents acting for users within the June figure; PPC Land reported that, within the bot majority, training-related crawlers alone accounted for 50.6% of all traffic at the time. Musk’s view that the 1,000x forecast is too low is his opinion, not a Cloudflare projection.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
Musk said agentic traffic already exceeds human traffic and that Cloudflare’s 1,000x five-year forecast is too low. Cloudflare’s Radar showed bots at 57.4% of web-content requests in June, about 1.35 times human requests by our arithmetic, and Prince has forecast automated traffic at 1,000 times human within five years.
A note on sourcing. Musk’s words are from our machine transcript of the posted clip. Cloudflare’s June Radar figures are as reported by SiliconANGLE and PPC Land; Prince’s posts are quoted from X and his 1,000x forecast as quoted by The Next Web from The Verge. Ratios are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
