OpenAI CEO Sam Altman has declared a “code purple” to focus firm sources on enhancing ChatGPT, in response to an inner memo reported by The Wall Avenue Journal and The Info.
The memo alerts OpenAI’s response to rising competitors from Google, whose Gemini 3 mannequin has outperformed ChatGPT in a number of benchmark checks since launching final month, in response to Google’s personal analysis information and third social gathering leaderboards.
What’s New
Altman instructed workers that ChatGPT’s everyday expertise wants enchancment. Particular areas embrace personalization options, response velocity and reliability, and the chatbot’s capacity to reply a wider vary of questions.
The corporate makes use of a color-coded system to point precedence ranges. This effort has been elevated to “code purple,” above the earlier “code orange” designation for ChatGPT enhancements.
A brand new reasoning mannequin is anticipated to launch subsequent week, in response to the memo, although OpenAI hasn’t publicly introduced it.
Delayed Merchandise
A number of product initiatives are being postponed because of this.
Promoting integration, which OpenAI had been testing in beta variations of the ChatGPT app, is now on maintain, in response to The Info. AI brokers designed for purchasing and healthcare are additionally delayed, together with enhancements to ChatGPT Pulse.
Altman has inspired momentary group transfers to help ChatGPT improvement and established every day requires these chargeable for enhancements.
Aggressive Context
On the technical aspect, Google’s Gemini 3 and associated fashions have posted robust scores on reasoning benchmarks. Google says Gemini 3 Deep Assume outperforms earlier variations on Humanity’s Final Examination, a frontier degree benchmark created by AI security researchers, and different tough checks. These outcomes are mirrored on Google’s personal Gemini 3 Professional benchmark web page and on impartial leaderboards that observe mannequin efficiency.
OpenAI hasn’t launched comparable public benchmark information for its subsequent reasoning mannequin but, so comparisons depend on present GPT 5 outcomes slightly than the upcoming system referenced within the memo.
Google can be persevering with to put money into generative picture instruments like its Nano Banana and Nano Banana Professional picture turbines, which sit alongside Gemini 3 as a part of a broader AI product lineup.
Benchmark Context
Humanity’s Final Examination is meant to be a more durable successor to saturated benchmarks like MMLU. It’s maintained by the Heart for AI Security and Scale AI, with an summary obtainable on the mission web site and outcomes tracked by a number of leaderboards, together with Scale’s official leaderboard and third social gathering dashboards similar to Synthetic Evaluation.
Google’s Gemini 3 Professional benchmark documentation lists the next rating on Humanity’s Final Examination than a number of competing fashions, together with GPT 5. That’s the idea for reporting that Gemini 3 has “outperformed” ChatGPT on that particular benchmark.
OpenAI has printed robust outcomes on different reasoning benchmarks for its GPT 5 collection, however the memo seems to be reacting to this current wave of Gemini 3 efficiency information slightly than a single check.
Visitors And Utilization Context
Regardless of the technical strain, OpenAI nonetheless has a big lead in assistant utilization.
In a current submit on LinkedIn, ChatGPT head Nick Turley mentioned ChatGPT is the “#1 AI assistant worldwide,” accounting for “round 70% of assistant utilization” and roughly “10% of search exercise.” You may learn his full feedback right here.
Separate reporting from retailers together with the Monetary Instances signifies OpenAI has greater than 800 million weekly customers, with most on the free tier, whereas Gemini’s consumer base has been rising shortly from a decrease start line.
Altman’s memo acknowledges Google’s current progress and warns of “momentary financial headwinds,” whereas additionally saying OpenAI is “catching up quick.”
A Acquainted Playbook
The “code purple” designation echoes Google’s personal response to ChatGPT a number of years in the past.
Google administration declared a “code purple” after ChatGPT’s viral launch. CEO Sundar Pichai redirected groups throughout Google Analysis, Belief and Security, and different departments to deal with AI product improvement.
That urgency led to the accelerated improvement of Google’s AI merchandise, culminating in Bard’s launch in early 2023 and its subsequent evolution into Gemini.
Now the roles have reversed. Google’s sustained funding in AI infrastructure has produced a mannequin that scores larger than ChatGPT on a number of excessive profile benchmarks, prompting OpenAI to undertake an analogous disaster response framework for its flagship product.
Firm Response
Nick Turley, OpenAI’s head of ChatGPT, addressed the aggressive panorama in current posts on LinkedIn and X, the place he described ChatGPT as the highest AI assistant worldwide.
“New merchandise are launching each week, which is nice,” he wrote in one of many posts, saying that competitors pushes OpenAI to maneuver quicker and proceed enhancing ChatGPT.
He added that OpenAI’s focus is making ChatGPT “extra succesful” whereas increasing entry and making it “extra intuitive and private.”
OpenAI hasn’t publicly commented on the leaked memo itself.
Trying Forward
OpenAI’s new reasoning mannequin launch will present the primary indication of how the corporate is executing on Altman’s directive. The delay of promoting and AI brokers suggests ChatGPT high quality has grow to be the corporate’s singular close to time period precedence, a minimum of internally.
For entrepreneurs and web optimization professionals, the extra fast affect is more likely to be on how ChatGPT handles complicated queries, analysis duties, and observe up questions as soon as the brand new mannequin is dwell. Any measurable modifications in reply high quality, velocity, or personalization can be vital to look at alongside Google’s continued Gemini 3 rollouts.
Featured Picture: Mijansk786/Shutterstock
