Credits: Village Global / Wikimedia Commons — CC BY 2.0.
In early December 2025, Sam Altman declared a “red alert” at OpenAI in San Francisco: focus all efforts on ChatGPT, as Google pushes Gemini 3.0 (including the Pro version) and the energy and financial bill is exploding. Projects put on hold, strategy shaken up, space ambitions lurking in the background. What does this refocusing mean for users, digital jobs, environmental impact, and AI governance?
What The “Red Alert” Memo Reveals
In early December 2025, an internal memo attributed to Sam Altman, CEO of OpenAI, circulated in San Francisco. It describes a state of red alert centered on ChatGPT, with a clear directive: re-center efforts on the company’s most visible consumer product. Teams are being asked to prioritize speed, reliability, personalization, and breadth of topics, at the expense of peripheral initiatives.
This document, reported by several media outlets, comes in a context described as “turbulent.” In the background: the upgrade of Gemini 3.0 at Google/Alphabet, seen internally as an immediate threat to ChatGPT’s appeal. Management argues that every delay, every unpolished feature, can push groups of users toward the competition.
Why Google Is Worrying OpenAI
According to this internal literature and press cross-checking, Gemini 3.0, especially the Pro version, appears to have exceeded certain expectations and benchmarks. The ecosystem is expanding, including on the video side with the Nano Banana tool. As a result, some advanced users are increasingly testing Google’s offering at ChatGPT’s expense.
On the scale front, Google claims about 650 million monthly users for Gemini, while OpenAI cites more than 800 million weekly users for ChatGPT. Two different metrics, but a catch-up dynamic is emerging. In a battle over adoption and habits, the slightest performance gap can carry a lot of weight.
OpenAI: Projects Suspended – Pulse, Ads, and Agents
The “red alert” comes with delays. The research assistant Pulse (OpenAI) is put on pause. The advertising formats being considered for ChatGPT and for GPTs are postponed. Other vertical agents (shopping, education, health, professional niches) see their timelines slip. The directive is clear: focus engineering on the core chatbot.
This choice delays the financial thesis that attributed around 20% of future revenue to advertising and secondary agents. The company emphasizes immediate user quality. As a result, it aims to stem churn and maximize retention. However, it is not hesitating to delay more experimental monetization levers.
OpenAI: 2025 Finances, A Strained Trajectory
According to estimates reported by specialized press, OpenAI is said to have posted record losses in the first six months of 2025. In addition, those losses would be on the order of $13.5 billion for $4.3 billion in revenue. And that, despite a valuation nearing $500 billion. Other analyses estimate OpenAI profitability pushed back to 2029.
Even more ambitious, data center plans pegged at more than $1,400 billion by 2030 are being discussed, involving cumulative fundraises of more than $200 billion. These amounts reflect infrastructure costs: cloud, chips (Nvidia, AMD), energy, and cooling. The stated strategy is one of hypergrowth financed by strategic partners. Those partners include Microsoft and SoftBank, as well as the markets.
Sam Altman and Stoke Space: The Space Angle
The other thread in the story reads like applied science fiction. In the summer of 2025, Sam Altman reportedly moved closer to Stoke Space, a Seattle startup building reusable rockets. The hypothesis under consideration: a majority stake, worth several billion dollars, to build a competitor to SpaceX. The talks have stalled, with no agreement at this stage.
Beyond the public rivalry with Elon Musk (OpenAI cofounder turned opponent, CEO of SpaceX and xAI), the strategic calculation would aim at a very practical goal: securing new compute capacity in the long run. Data centers could follow the rockets.
Data Centers In Orbit: Solution Or Mirage?
The idea of space data centers has been circulating in tech literature for years. In orbit, cooling would benefit from a more stable environment. In addition, solar power generation would be abundant. Some latency could also be optimized for specific use cases. On the other hand, there are launch costs to consider. Likewise, maintenance constraints and space debris would pose a problem. Finally, questions of sovereignty and space law arise.
On the environmental side, the argument for a potentially better water and energy footprint deserves independent proof. Transporting heavy equipment into orbit comes with a non-negligible upfront carbon cost. The debate remains open: a future solution for lean compute, or a story meant to reassure investors?
Risks, Liability, and Procedures
On November 6, 2025, seven plaintiffs brought a case before the California courts for manslaughter. In addition, the accusation of assisted suicide was made after the deaths of four users. This death is attributed to the use of GPT-4o. These are allegations made as part of an ongoing proceeding. At this stage, no court ruling has confirmed these claims.
In the internal memo as reported, no explicit new measure is mentioned for vulnerable audiences. This point will probably prompt political pushback and a debate over the safety of these systems. In addition, suicide risk prevention will be discussed, as will the duty of care of platforms.
What This Changes For The Public
For users, the immediate effect could be positive: a ChatGPT that is faster, more reliable, more customizable, and more accessible to the broadest possible audience. But the ecosystem is becoming concentrated around a few players with an unmatched scale advantage. Access to high-performing AI assistants is becoming a consumer good. It is also, in practice, a public service issue.
For digital employment, the assistant battle is accelerating demand for orchestration skills (advanced prompting, integration, quality control) and for monitoring of drift (hallucinations, bias, safety). On the energy side, the trajectory of data centers and chips promises rising consumption, hence the pressure for real efficiencies here; orbit may not be the miracle solution.
Act 1: The Alert In The Hallways
In OpenAI’s hallways, you can imagine the dashboards turning orange: tickets waiting, latencies stretching, feedback piling up. The “red alert” sets the tempo. Cut what distracts, polish the core experience. In this endurance race, the first meter matters just as much as the last.
Act 2: The Billionaire Who Dreamed Of Rockets
In Seattle, Stoke Space dreams of circular engines and second stages that come back to land. Sam Altman is looking further ahead: compute in orbit, a network above the clouds, cooling without fresh water. The plan has the madness of the pioneers and the arithmetic of Excel spreadsheets. Between vision and loss of focus, the line is thin.
Act 3: What’s Left?
What remains is the question that obsesses investors and regulators: can OpenAI hold on until 2029 without profitability? Does the bet on massive fundraises, the idea of data centers valued in billions of dollars, and the product pauses add up to a credible path? Or are we watching the story of a bubble looking for its balance?
In a democracy, the answer cannot be left to the markets alone. It calls for guardrails: transparency on energy costs, independent assessment of risks, interoperability and portability to avoid lock-in to platforms, and active protection for vulnerable users.
Open Questions
- Can OpenAI finance its trajectory through 2030 without breaking its main product?
- Do space data centers amount to serious engineering or to storytelling for sovereign wealth funds and major investors?
- Does this assistant war benefit the public (quality, price, rights), or does it increase our dependence on a few platforms?
- Where are the additional protections for fragile audiences in this permanent state of urgency?
Methodological note: this article relies on internal documents reported by the press and on analyst estimates. The lawsuits mentioned concern ongoing proceedings. As such, the alleged facts cannot be considered established. In addition, this remains true as long as no final court decision is handed down.