• 🚨 OpenAI hits “Code Red” and pushes ads to the back burner – The Information reports that Sam Altman has declared an internal “code red” to prioritise improving ChatGPT as pressure intensifies from Google and other AI competitors.
  
  OpenAI will pause or slow several initiatives – including ads, agents and Pulse – to redirect resources into ChatGPT performance, reliability and user experience. The memo suggests a tactical refocus - prioritising defence of ChatGPT’s lead over new expansion, at least for now.
  
  

  Key takeouts:
  An extremely eventful couple of weeks at OpenAI has led to Sam Altman declaring a 'code red' as Google, with a focus on Gemini 3, has closed the gap in the AI race, following ChatGPT's clear lead.
  
  Altman has been accused of lacking focus and discipline when it comes to the vision of OpenAI and where to allocate costs and resources. He has a habit of spinning up enormous ideas – ChatGPT agents, AI hardware with Ive, a nuclear-powered compute strategy, crypto identity via Worldcoin, even custom silicon – often all at once. Supporters say it reflects ambition whereas his critics call it a lack of focus.
  
  Everything has seemed a bit rushed. Atlas was launched just in time for the holiday season and it felt unfinished. Yes, important to hit deadlines and get the product out into real users hands, but the experience feels clunky, the search tab feels half-cooked and it will take a lot for me to drop Chrome as a priority with Comet second.
  
  2026 is a huge year for Altman, OpenAI and ChatGPT. With a potential IPO on the horizon, ambitious subscriber growth targets, and pressure rising from Anthropic and Google, Altman faces a monumental challenge to prove investors have backed the right horse.

• 💰 Leak confirms OpenAI is preparing ads inside ChatGPT – BleepingComputer reports that strings in the latest ChatGPT Android beta reference an “ads feature”, “bazaar content”, “search ad” and a “search ads carousel”.
  
  The article notes that ChatGPT is now handling hundreds of millions of users a week and billions of prompts a day, putting it in the same league as major consumer platforms. With that kind of attention and intent data, the economics for personalised, conversation-native advertising are obvious – even if the first phase is limited to search-style answers.
  
  

  Key takeout: This story was doing the rounds on LinkedIn, but for regular readers of this newsletter, this comes as no surprise.
  
  Earlier this year, OpenAI forecast significant growth year-on-year from 2026 to 2030, which was reported by The Information.
  
  Right now, there isn’t an incentive to give this information to marketers, but if OpenAI wants to replicate what Google did in search and Facebook did in social, then they will build a full suite of ad-tech tools.
  
  Given OpenAI’s recent “code red” (see above), the leaked memo suggests that ads will be delayed – but make no mistake, ads will be coming to ChatGPT, it’s just a matter of “when”.
  
  OpenAI’s Head of ChatGPT, Nick Turley, has recently been vocal about zero public testing on ads when users thought they saw ads in their ChatGPT answers (the one that I saw looked more like an integrated app than an ad).
  
  Along with ads, Turley’s biggest priorities include personalisation, connecting ChatGPT to other apps such as email and Slack, and boosting its search capabilities — notably enabling ChatGPT to take actions for users.
  
  This was a discussion I had with Christian J Ward, Chief Data Officer at Yext. Watch the sneak peek on how Christian believes ChatGPT could evolve to include a secondary ad pane.

Andy

• 🎯 GEO vs SEO - is it evolution, rebranding, or just marketing theatre?
• 🧠 How users behave in AI search - longer prompts, multi-turn discovery, and Time-to-Result thinking.
• 🔍 Schema, llms.txt and cat.txt - what’s real, what’s placebo, and where evidence actually exists.
• 🧪 The Google exploit - how Mark’s team uncovered site quality score data & why it matters.
• 🚀 AlsoAsked - the origin story & SaaS reality - MRR growth, plateaus, churn and product pain points.