The Media Copilot

Blocking the bots is step one. People Inc.'s Jonathan Roberts on what comes next: the companies reselling publisher content, who sets the price, and why AI is spending its money in the wrong place.Pete Pachal talks with Jonathan Roberts, chief innovation officer at People Inc. (People, InStyle, Investopedia), about blocking every bot by default and allowing 38 through, the disguises blocked bots wear, why robots.txt is a chocolate teapot, the "picks and shovels" companies taking content for free and selling it for money, the 18 Napsters of content with no Spotify, serving AI companies Markdown instead of HTML, who gets to set the price, per crawl versus per use, why models without reference are fiction writers, why inputs now matter more than the model, TIME and Mobian's ads for bots, and why automating the chore should never mean automating the choice.Chapters0:00 Why the demand for information moved upstream2:30 Jonathan Roberts, welcome back to the Media Copilot2:36 Can smaller publishers really block bots?4:29 Yes, and the future is hungrier for content than ever6:47 Why robots.txt is a chocolate teapot8:17 The CDNs are the new front line9:48 Block everything, allow 3810:30 Blocked bots put on disguises12:11 Botnets, hijacked smart TVs and the FBI13:38 What a smaller publisher can do today17:02 Chips, power and information19:21 18 Napsters and no Spotify22:14 A better product for bots than HTML25:45 Who gets to set the price?28:40 Per use or per crawl?32:05 Information is a supply chain, not a mine34:54 Training, inference and fiction writers39:41 TIME, Mobian and ads for bots42:49 Automate the chore, not the choice43:24 Search is humble, chatbots are arrogant47:03 People Inc.'s LLM visibility program48:12 Two futures for publishers50:39 OutroGuestJonathan Roberts, Chief Innovation Officer, People Inc.People Inc.MentionedPublishers are fighting the wrong battle against AI (Pete's column)Publishers finally have AI buyers but no say in the priceWhat TIME learned from a month of selling ads to bots, with Mark HowardJonathan's first episode, April 2025Watch on YouTube | Full post and transcript | Subscribe on SubstackProduced by The Media Copilot.Music: "Favorite" by Alexander Nakarada, licensed under CC BY 4.0

When AI answers the question, publishers lose more than the click. Bombora's Jaime Schultheis on why AI referrals convert, why Google is a shaky foundation, and what AI companies should give publishers besides a check.Pete Pachal talks with Jaime Schultheis, head of global data partnerships at Bombora, about what Bombora's publisher data shows (steeper referral declines, more bots, niche B2B titles holding up), why AI referral traffic is small but converts far better than search, Google's new option to opt out of AI features, events and communities as "AI proof" experiences, measuring the halo effect for sponsors, sorting good bots from bad, TIME and Mobian's ads-for-bots model and Perplexity's pushback, and why she's optimistic that publishers will come out of this with more leverage.Chapters0:00 What a website visit is really worth1:59 Jaime Schultheis, welcome to the Media Copilot2:04 From newspapers to Bombora3:39 "I probably was not worried enough"4:53 What Bombora's publisher data is showing6:17 Why specialty publishers are holding up9:42 What's the reward for GEO?12:10 Charging more for AI referral traffic14:11 Don't build your business on Google15:25 Google's new AI opt-out16:59 Events, communities and "AI proof" experiences20:37 The first step for a small publisher23:13 Measuring the halo effect for sponsors26:09 Getting into events without the pitfalls28:58 Good bots, bad bots32:35 What a fair AI deal should give back37:05 TIME, Mobian and Perplexity's pushback41:34 What worries her, and why she's optimistic43:52 OutroGuestJaime Schultheis, Head of Global Data Partnerships, BomboraLinkedIn | BomboraMentionedThe Atlantic's AI Gamble with Nicholas ThompsonWhat TIME learned selling ads to bots, with Mark HowardWatch on YouTube | Full post and transcript | Subscribe on SubstackProduced by The Media Copilot.Music: "Favorite" by Alexander Nakarada, licensed under CC BY 4.0

Rob Kelly tracks 99 AI licensing deals. What publishers are really selling now, and why OpenAI may settle with the Times.Kelly, the creator and host of Media & the Machine, keeps a spreadsheet of 99 AI content licensing deals, from The New York Times' little-discussed $100 million Google arrangement to Yelp's new deal with OpenAI. He ran the content licensing business at CMP Media, so he knows how profitable a licensing feed can be. Pete Pachal and Kelly cover the two early playbooks (the Times sues OpenAI, News Corp signs with it), why training is fading from the deals as inference and retrieval catch up, what Kelly calls licensing content to commerce, the squeeze on mid-tier publishers, and his prediction that OpenAI settles with the Times before its IPO.In this episodeKelly's path from Wall Street and tech journalism to running content licensing at CMP MediaThe two early playbooks: the Times sues OpenAI, News Corp signs a $250 million dealThe Times' little-discussed $100 million licensing deal with GoogleWhy licensing revenue is close to pure profit for publishersHow deals have shifted from training toward inference and RAG, and why the models haven't run out of training dataWhether AI companies pay mainly to avoid lawsuitsLicensing content to commerce: Yelp's request-a-quote button and a subscribe button inside the answerGoogle Zero and taking a beginner's mind to AI trafficHow the AI answer is being renegotiated, from ads for bots to brand linksThe mid-tier squeeze, ProRata, and the RSL collective licensing protocolWhether Google or Apple could import the YouTube revenue-share modelThe backlash against AI among younger people, and an 8.7x productivity gain measured in a spreadsheetTwo futures for media, and why Kelly expects OpenAI to settle with the Times before its IPOGuest: Rob Kelly, Creator and Host, Media & the Machine: LinkedIn | Media & the MachineWatch the video version: YouTubeFull post and transcript: mediacopilot.aiAbout the show: To explore more conversations like this and see what's new, visit the Media Copilot website at mediacopilot.ai. You'll find new episodes, expanded resources, and tools designed for journalists, communicators, and media leaders navigating the fast-changing world of AI.Subscribe to The Media Copilot on Substack, Apple Podcasts, Spotify, or your favorite app. On YouTube? Like and subscribe: youtube.com/@themediacopilotProduced by The Media Copilot.Music: "Favorite" by Alexander Nakarada, licensed under CC BY 4.0All rights reserved. © AnyWho Media 2026

Bot traffic has already passed human traffic on the web. Oasy co-founder Choy Travers thinks publishers should stop blocking the bots and start selling them ads.Pete Pachal talks with Choy Travers, co-founder of Amsterdam-based Oasy, about a new kind of advertising: sponsored content served to AI crawlers on a publisher's site, on the bet that the AI will use it when answering a user's question. They cover why blocking bots backfires, how Oasy's free-for-publishers, CPM-for-advertisers model works, disclosure that has to survive a crawl and a summary, Perplexity's cloaking objection, why licensing deals leave most publishers behind, and the billboard problem in AI attribution. This is part two of a two-part look at ads for bots; part one is Mark Howard, COO of TIME.Disclosure: Pete has been running Oasy on mediacopilot.ai for several months.In this episodeWhy blocking AI crawlers means removing yourself from the fastest-growing search channelHow Oasy's model works: free for publishers, CPM for advertisers, one placement per pageWhat an advertiser is actually buying when there's no guaranteed placementDisclosure that has to survive a crawl, a chunk, and a summaryPerplexity's cloaking objection, and what it gets right and wrongWhy licensing deals will pay the top players and leave the rest scrapedWho is responsible when an AI platform strips the labelPay-per-crawl versus pay-per-use, and the billboard problem in attributionThe utopian and dystopian versions of the ads-for-bots futureGuest: Choy Travers, Co-Founder, Oasy: LinkedIn | oasy.aiWatch the video version: https://www.youtube.com/watch?v=nU43Zhz3esAFull post and transcript: mediacopilot.aiAbout the show: To explore more conversations like this and see what's new, visit the Media Copilot website at mediacopilot.ai. You'll find new episodes, expanded resources, and tools designed for journalists, communicators, and media leaders navigating the fast-changing world of AI.Subscribe to The Media Copilot on Substack, Apple Podcasts, Spotify, or your favorite app. On YouTube? Like and subscribe: youtube.com/@themediacopilotProduced by The Media Copilot.Music: "Favorite" by Alexander Nakarada, licensed under CC BY 4.0All rights reserved. © AnyWho Media 2026

Bot traffic passed human traffic on the web this year. TIME decided to sell the bots ads. One month in, COO Mark Howard explains what advertisers are buying, why the AI seems to remember what it scrapes, and what Perplexity got wrong.Pete Pachal talks with Mark Howard, chief operating officer of TIME, about agent ads: brand-verified facts, labeled as sponsored, placed on the machine-readable versions of TIME pages for the AI crawlers it allows in. They cover the infrastructure TIME built first (TollBit, ScalePost, a 70-agent allow list), the Mobian brand-gap method behind each campaign, the two measurement studies, the referrals-versus-citations puzzle, what an advertiser actually buys, whether the models can read the disclosure, the Perplexity dispute, what counts as a legitimate crawler, and why pay-per-crawl is both Howard's biggest worry and biggest hope. This is part one of a two-part look at ads for bots; part two is Oasy co-founder Choy Travers.In this episodeWhy TIME chose to learn by doing instead of waiting for the market to settleThe infrastructure underneath agent ads: TollBit, ScalePost, and a 70-agent allow listThe markdown page built for machines, and what an agent ad actually isWhich bots show up, and what they're there forThe TIME100 Creators moment that made the case for the productMobian's method: visibility, favorability, and accuracy across the major modelsWhy TIME believes the agents remember what they scrapeOne month of data: causal lift and model impact over timeWhat the advertiser is buying when there's no guaranteed placementWhether the models can read the sponsored label, and what they do with itThe Perplexity dispute, and what OpenAI thinksWhat counts as a legitimate crawlerPay-per-crawl as both the biggest worry and the biggest hopeGuest: Mark Howard, Chief Operating Officer, TIME: LinkedIn | time.comWatch the video version: https://www.youtube.com/watch?v=rjz-GEaibbIFull post and transcript: mediacopilot.aiAbout the show: To explore more conversations like this and see what's new, visit the Media Copilot website at mediacopilot.ai. You'll find new episodes, expanded resources, and tools designed for journalists, communicators, and media leaders navigating the fast-changing world of AI.Subscribe to The Media Copilot on Substack, Apple Podcasts, Spotify, or your favorite app. On YouTube? Like and subscribe: youtube.com/@themediacopilotProduced by The Media Copilot.Music: "Favorite" by Alexander Nakarada, licensed under CC BY 4.0All rights reserved. © AnyWho Media 2026