OpenAI doesn’t wish to be omitted. The week after Anthropic introduced Claude Mythos Preview—an AI mannequin that has put governments all over the world on edge due to its potential means to hack into banks, vitality grids, and navy techniques—OpenAI shared a program that’s uncannily related. And similar to Anthropic did with its mannequin, OpenAI has, for cybersecurity functions, restricted entry to this new bot, referred to as GPT-5.4-Cyber, to a small group of trusted customers.
This sequence has turn out to be one thing of a sample: First Anthropic will make an announcement, after which OpenAI will observe swimsuit. Final 12 months, Anthropic launched Claude Code, an AI coding instrument. A few months later, OpenAI got here out with its personal model, Codex. When Claude Code had a breakout second in January, OpenAI responded with two main updates to Codex alongside a press blitz for the product. And earlier this month, OpenAI launched a model of Codex that enables it to make use of different apps in your desktop—much like an present Anthropic instrument referred to as Claude Cowork.
Till lately, Anthropic—based by a bunch of former OpenAI workers in 2021—performed the function of youthful brother. OpenAI kicked off the whole AI growth with the discharge of ChatGPT, and has had extra customers, funding, and title recognition ever since. However Anthropic has been driving excessive on the explosive reputation of Claude Code and booming gross sales of its AI fashions to giant firms. The agency’s showdown with the Pentagon has additionally helped vault it into the general public eye. In early April, Anthropic stated its income price had hit $30 billion a 12 months—showing to surpass OpenAI’s.
In its public messaging, OpenAI has been detached and even considerably derogatory towards Anthropic. Final week, when OpenAI launched its latest mannequin, GPT-5.5, the announcement was paired with direct and veiled references to the way it beat out Anthropic’s newest, Claude Opus 4.7. However internally, the agency is seemingly on edge. In a current leaked company-wide memo, Denise Dresser, OpenAI’s chief income officer, felt the necessity to handle one explicit competitor: “Right here are some things price conserving in thoughts, particularly on Anthropic.” The rival agency’s product choices are slender, Dresser wrote, and “their story is constructed on concern,” referencing Anthropic’s loud messaging concerning the risks of AI. “Our constructive message will win over time.” (OpenAI, which has a enterprise partnership with The Atlantic, didn’t reply to a request for remark. Anthropic additionally didn’t reply to a request for remark.)
If imitation is the sincerest type of flattery, OpenAI’s actions are particularly telling. At each flip, OpenAI has appeared keen to repeat the success of its rival. For starters, as Anthropic’s specific deal with mitigating the dangers of AI has apparently gained the belief of many shoppers, OpenAI has imitated a lot of its rival’s security initiatives. In early 2026, after Anthropic revealed a significant replace to Claude’s “Structure,” a doc that tells the AI mannequin methods to behave, OpenAI launched a significant marketing campaign round its equal doc.
However OpenAI’s most essential, Anthropic-esque pivot has been in its enterprise mannequin. Early on, these two firms made essentially completely different bets on how they might finally make cash. OpenAI positioned itself as a client behemoth, hoping to capitalize on ChatGPT’s a whole lot of thousands and thousands of customers. Final fall, the corporate launched the AI-video app Sora and an AI-powered internet browser. OpenAI has made forays into e-commerce and is testing advertisements in ChatGPT. Once in a while, the corporate teases the AI gadget that it’s creating with the previous Apple designer Jony Ive. Anthropic, in the meantime, has targeted on the much less flashy purpose of promoting its AI instruments to companies and software program engineers.
Regardless of OpenAI’s quite a few benefits, Anthropic’s deal with code and enterprise prospects appears to be profitable. Though OpenAI is price extra based mostly on the latest fundraising rounds, Anthropic now has a greater valuation—greater than $1 trillion—in some personal markets. Anthropic’s explosive development is especially essential as the 2 firms each race to go public, in flip accessing an enormous pool of latest traders, and attempt to show they’ll finally be worthwhile. (Each firms nonetheless have an extended method to go in that regard.)
OpenAI is now desirous to catch up. In December, OpenAI employed Dresser, a former CEO of Slack, to pursue extra enterprise prospects. In late January, Altman gathered a number of main executives for a lavish dinner in San Francisco to preview all the enterprise choices his firm was planning, in accordance to The Data. The corporate has since made a blitz of bulletins round coding instruments and enterprise AI choices, together with a brand new set of “Frontier Alliances”: partnerships with a number of of the world’s premier consulting companies, together with McKinsey & Firm and Boston Consulting Group, to speed up enterprise adoption of ChatGPT. In mid-March, one other inside OpenAI memo reportedly acknowledged that the corporate wanted to remove “aspect quests” and deal with the enterprise and coding markets. Anthropic’s success in these areas, the memo acknowledged, needs to be a “wake-up name” for OpenAI. The agency additionally scrapped Sora and has been aggressively promoting and messaging about Codex for months now. “I’m joyful everyone seems to be switching to Codex,” Altman wrote on X earlier this month.
OpenAI’s pivot to its enterprise enterprise has not been whole. It did, as an illustration, lately shell out reportedly a whole lot of thousands and thousands of {dollars} to amass a distinct segment tech podcast. And Anthropic, for its half, has needed to take some cues from OpenAI—notably by making large and costly data-center offers, similar to an growth in its partnership with Amazon Net Companies. Anthropic’s CEO, Dario Amodei, has beforehand insinuated that OpenAI has made such offers “as a result of it sounds cool.”
Which firm will win the AI race is anyone’s guess. Regardless, OpenAI’s embrace of the Anthropic enterprise mannequin makes one factor abundantly clear: For all of the marvel and alter that generative AI brings as a expertise, there hasn’t been any actual innovation within the enterprise fashions of Silicon Valley. For many years, most tech firms have succeeded by both promoting advertisements (the route of Meta and Google) or promoting enterprise instruments (like Salesforce and Slack). At some point OpenAI or Anthropic would possibly treatment most cancers and remake the world, however for now they nonetheless must pay the payments.
