Sam Altman Declares “Code Red” as OpenAI Confronts Rising Pressure from Gemini 3 and Big Tech Rivals

• Internal “code red” memo warns of critical moment for ChatGPT (December 2025) Sam Altman sent an internal memo stating “We are at a critical time for ChatGPT,” initiating daily calls for teams responsible for core improvements and encouraging temporary employee transfers to increase development velocity. The message notes that ChatGPT needs better speed, reliability, personalization, and day-to-day user experience. The directive delays work on advertising integration, health and shopping AI agents, and a personal assistant tool called Pulse. It also echoes Google’s own internal “code red” declaration from December 2022, when Google scrambled to respond to ChatGPT’s early explosion in popularity. The new push comes three years after ChatGPT’s launch and signals a return to emergency footing.
• Google’s Gemini 3 surpasses rivals on major AI benchmarks Gemini 3, released in mid-November 2025, quickly topped LM Arena’s crowdsourced leaderboard and gained praise for improvements in reasoning, speed, multimodal fluency, and video processing. The model’s rapid performance leap intensified competitive pressure on OpenAI. Google’s AI growth is reinforced by its lucrative search business, allowing heavy reinvestment into compute, model training, and deployment infrastructure.
• High-profile industry figures publicly switch to Gemini 3 Salesforce CEO Marc Benioff announced on X that after testing Gemini 3 for two hours, he would not return to ChatGPT after using it daily for three years. He described Gemini 3’s leap as “insane,” praising performance across reasoning, imagery, and speed. His comments accelerated momentum around Gemini 3’s perceived superiority and shaped broader industry sentiment.
• OpenAI delays its ChatGPT advertising rollout to focus on quality gains OpenAI postponed advertising integration, additional AI agents, and the Pulse assistant to redirect resources to core improvements. The shift comes as ChatGPT exceeds 800 million weekly users, while Google’s Gemini app grows rapidly — from 450 million monthly active users in July to 650 million in October, based on Business Insider data. OpenAI also introduced the Atlas browser in October to compete with Google Chrome as users increasingly rely on AI for real-time answers. Despite these expansions, OpenAI has not begun monetizing ads within ChatGPT, keeping its revenue centered on premium subscriptions.
• Leadership signals long-term expansion goals despite external pressures Nick Turley, head of ChatGPT, emphasized that online search represents one of ChatGPT’s largest opportunities as the company focuses on making the product more capable, intuitive, and personal. His comments align with the internal “code red” directive and reflect a broader attempt to push ChatGPT into areas dominated by search engines. Turley’s messaging coincides with ChatGPT’s third anniversary and a renewed commitment to capability growth across global markets.
• Massive valuation and revenue ambitions amid ongoing losses OpenAI’s valuation has reached $500bn, up from $157bn in October 2024. Although unprofitable, OpenAI expects over $20bn in annual revenue by year-end and aims to grow into “hundreds of billions” by 2030. Reporting indicates that OpenAI holds more than $1 trillion in financial obligations to cloud providers and chip manufacturers to secure compute capacity — commitments that sit alongside its planned $1.4tn datacenter investment over the next eight years. These obligations heighten the financial stakes, with analysts warning that OpenAI is stretching itself across too many fronts as investor concern about an AI-market bubble grows.
• Strategic funding from Microsoft and SoftBank provides runway Microsoft and SoftBank continue to support OpenAI’s capital needs, sustaining the company’s ability to execute its multi-year, high-compute vision. This funding is crucial given OpenAI’s lack of profitability and its immense infrastructure commitments. Unlike Google and other rivals whose AI operations are subsidized by ad revenue or retail businesses, OpenAI relies heavily on fundraising, partnerships, and premium subscriptions.
• Altman highlights compute scarcity as the primary operational risk Altman emphasized that insufficient compute availability remains the company’s biggest constraint, outweighing concerns about overspending. He noted intensifying user demand, expanded multimodal workloads, and rapid competitive iteration as factors stressing infrastructure. His memo referenced a new simulated reasoning model expected to release next week, which may outperform Gemini 3 in internal tests. The cycle of one-upmanship continues as long as funding and compute capacity allow.
• Apple reorganizes its AI leadership to stay competitive Apple replaced John Giannandrea with Amar Subramanya as its new AI lead, citing the need to restructure AI initiatives across foundation models, machine learning research, and AI safety and evaluation. The remaining parts of the organization were reassigned to Sabih Khan and Eddy Cue to streamline alignment. Apple acknowledged delays in Siri’s major AI upgrade to 2026. The restructuring comes amid morale challenges, repeated staff losses to competitors, and a shorthanded AI division that will require reinforcements to recover momentum. These internal dynamics heighten pressure as OpenAI and Google accelerate product improvements.
• Samsung and other rivals move faster than Apple on AI integration Samsung continues to deploy device-level AI advancements at a faster pace than Apple. As OpenAI and Google escalate the race with a “code red” posture, expectations for Apple’s next-generation Siri have increased. Apple’s slower timeline magnifies competitive pressure, with its upcoming AI deployments needing to meet a significantly raised industry standard by the time they launch.

Why This Matters

The rapid ascent of Gemini 3, the accelerating adoption of Google’s AI products, Apple’s urgent internal restructuring, and OpenAI’s expanding financial exposure have created one of the most competitive inflection points in AI since ChatGPT’s original launch. High-profile user shifts, trillion-dollar compute obligations, and delays to OpenAI’s product roadmap underscore a shifting power balance. Leadership in AI is increasingly determined by compute scale, capital endurance, organizational flexibility, and the ability to adapt quickly in a volatile, rapidly escalating environment.

This article was drafted with the assistance of generative AI. All facts and details were reviewed and confirmed by an editor prior to publication.