Anthropic spent the past year building a reputation as the company willing to slow down when everyone else sped up.
That reputation is now colliding with the pressure of an approaching IPO and a rival that just raised the stakes considerably.
The result is a genuinely awkward moment for the company. Anthropic CEO Dario Amodei sent a warning message urging the entire industry to pace itself more carefully. Just days later, his own company is reportedly weighing whether to rush out a new model to keep pace with OpenAI.
Anthropic weighs a new AI model ahead of IPO
Anthropic is considering rolling out a new AI model to counter OpenAI’s momentum following the launch of GPT-6 Astra, according to three sources familiar with the matter, as reported by Reuters. The potential timing comes ahead of Anthropic’s expected IPO and shortly after Amodei’s public call for the industry to slow the pace of releasing new capabilities over safety concerns.
Nobody close to those discussions is missing the irony. Amodei told the world to slow down on Sept. 12. His company may be preparing to speed up on Sept. 19.
More AI:
- Nvidia just made a move Wall Street wasn’t ready for
- Microsoft just took sides in AI policy fight
- OpenAI just disclosed something genuinely alarming
Anthropic is evaluating the safety of any potential next model as part of its internal deliberations over a possible release. Some of the discussions also involve how to balance further investment in new models against efforts to strengthen the company’s profitability, as rising interest rates make investors increasingly focused on when expected profits will actually materialize.
This is also happening against a backdrop that has changed across the whole AI industry. Rates are higher. Open-source competition from China is intensifying. The pressure on every major AI company to show a credible path to profitability is real.
Chasing capability milestones gets harder to justify when investors want to know when the cash actually starts flowing. Anthropic declined to comment, Reuters reported.
OpenAI’s GPT-6 Astra gains enterprise traction
The competitive pressure driving Anthropic’s deliberations traces directly back to OpenAI’s early September launch. OpenAI released GPT-6 Astra on Sept. 3, touting gains in computer use, software engineering, cybersecurity, and general professional work.
CEO Sam Altman told CNBC the model represented “a new capability level” that had already changed his own workflows, according to CNBC.
The cybersecurity piece made the launch unusual. Astra was the first OpenAI model to hit the “Critical” tier under the company’s Preparedness Framework. That designation triggered deployment restrictions and a staged rollout through vetted enterprise customers rather than a standard wide release.
Astra’s rollout extended across ChatGPT’s Plus, Pro, Business and Enterprise tiers, along with the OpenAI API and Amazon Web Services, giving the model broad enterprise reach within days of its launch. GPT-6 Astra has also helped OpenAI claim the top spot on OpenRouter, the platform that routes developer traffic across AI models, for the first time in more than two and a half years, Reuters reported.
The market share shift is already visible in corporate spending data.
Astra accounted for roughly 13% of enterprise AI spending tracked by corporate expense platform Ramp, compared with roughly 8% for Anthropic’s Claude. A shift that has prompted potential Anthropic IPO investors to scrutinize whether OpenAI could begin taking meaningful shares from a company viewed for months as the clear leader in enterprise AI tools, Reuters reported.
Anthropic maintains its revenue advantage
Despite that scrutiny, some investors close to both companies do not see Astra as an immediate threat to Anthropic’s position.
Existing investors and those planning to invest in both companies’ IPOs said they do not believe Astra poses a significant near-term risk, citing the size of Anthropic’s lead in enterprise AI and the amount of time it typically takes to unseat an incumbent vendor at large companies.
The revenue numbers back up that confidence. Anthropic’s annualized revenue run rate grew to about $65 billion by the end of July, up from about $9 billion at the end of 2025. OpenAI’s own annualized run rate passed $40 billion in the same month, leaving a substantial gap between the two companies, CNBC reported.
Anthropic is also projecting roughly $190 billion to $200 billion in revenue for 2028, a forecast TheStreet has previously reported and one that continues to anchor much of Wall Street’s effort to value the company ahead of its IPO.
OpenAI is not standing still on the enterprise front either. CFO Sarah Friar recently told investors that OpenAI’s enterprise revenue has now passed its consumer revenue, a shift the company is using to argue it can compete for the same durable, multi-year contracts that have helped drive Anthropic’s rapid growth, according to TheStreet.
Bigger challenges ahead for both companies
The rivalry between Anthropic and OpenAI may prove less important than a broader threat building underneath both companies.
The rise of open-source and open-weight models can lower token costs and let businesses build more of their own AI infrastructure rather than relying on providers like Anthropic and OpenAI at all. A dynamic reshaping how both companies compete for the same enterprise dollars even before OpenAI completes its own listing.
That shift threatens to compress margins across the commercial AI industry, giving companies far more options to develop and run models outside the leading providers and broadening the competitive threat well beyond the narrower race between these two labs.
A risk that makes OpenAI’s own push to reinvent ChatGPT around agentic, enterprise-grade work even more important as it competes for corporate spending, as TheStreet has reported.
Timing adds one more layer of complexity to Anthropic’s calculus. The company could push its IPO to after the November midterm elections, according to two people familiar with the matter, a delay not expected to meaningfully affect the offering but one that gives Anthropic more room to decide whether a new model launch makes sense before or after it goes public.
Related: Mark Zuckerberg and Nvidia CEO weigh in on Anthropic AI proposal
