Super Bowl 60 is this weekend, and the New England Patriots and Seattle Seahawks are not the only ones battling for a title.
Generative AI startups and bitter rivals OpenAI and Anthropic are using the premier U.S. sports event to release ads and promote products, as both model makers endeavor to win over consumers and enterprises. However, while Anthropic’s ads are entertaining, the winner of the AI race won’t be the vendor that puts out the most targeted ads, but the one that provides the best value. Although OpenAI is expected to release a 60-second ad on Feb 8. Super Bowl Sunday, Anthropic this week previewed four different Super Bowl commercials (OpenAI has not previewed its commercial).
All the Anthropic spots have the same premise. A human character is sharing or asking a question to a chatbot-like character. The chatbot starts answering the question, then pivots to promoting a product or an ad. Each commercial ends with the same tagline: “Ads are coming to AI. But not to Claude.”
That Claude ad tagline is a direct shot at its adversary, a humorously barbed response to the ChatGPT maker’s declaration on Jan. 16 that it would start testing ads in the free version of its globally popular AI chatbot, after previously saying the move would be a “last resort” in 2024.
Anthropic’s commercial is the latest salvo in the high-pitched AI battle between the two frontier model makers, highlighting the increasingly tense push-and-pull in their fight to become the premium AI provider of choice for enterprises and consumers.
While OpenAI has been the leader in the consumer market, with 800 million to 900 million active ChatGPT users each week, Anthropic has carved out a reputation as the gold standard in the enterprise, and it has momentum. The vendor, founded by former OpenAI executives and researchers, has positioned itself as a responsible AI model maker with its Constitutional AI framework and Model Context Protocol generative AI standard, making it a preferred choice in enterprises over OpenAI. At least for now. Google, which competes with independent generative AI vendors, has also seen notable progress with its Gemini model family, for both businesses and consumers.
Two Different Customers
“We’re talking about consumer versus enterprise,” said David Nicholson, an analyst at Futurum Group.
But both vendors are also trying to gain ground in markets where the other vendor dominates. Anthropic is trying to make a play in the consumer arena, while OpenAI is trying to show it is enterprise-ready, with a full stable of models for business use.
“They are going after each other for the areas that they are trying to get into, and Anthropic is saying, ‘We’re all enterprise, we’re not going to sell your data for ads,'” said Ray “R” Wang, CEO and founder of Constellation Research.
In response to Anthropic’s commercials, OpenAI’s CEO Sam Altman posted on X that while the ads are “funny,” they are also “dishonest.”
“Anthropic serves an expensive product to rich people,” Altman wrote. “We are doing that too, but we also feel strongly that we need to bring AI to billions of people who can’t pay for subscriptions.”
Altman added that more Texans use ChatGPT for free than use Claude in the U.S., an assertion that is likely all but unverifiable.
The battle has seen nearly daily skirmishes in the run-up to the Super Bowl, which has for decades served as a stage for the most spectacular and ingenious ads by major companies.
OpenAI also released its Frontier model on Feb. 5, which competes directly with Anthropic’s Claude Cowork AI agent. That product release sparked a panicky market sell-off of legal information providers’ stocks earlier this week, when Anthropic launched a series of plugins for vertical industries, including legal, triggering fear among traditional software suppliers in certain sectors. Meanwhile, the Frontier platform helps enterprises build, deploy, and manage AI agents to function as AI coworkers. A few hours later on Feb. 5, Anthropic introduced Opus 4.6, an upgrade of its popular advanced reasoning model. The model supports a massive million-token context window, enabling it to process codebases and large legal and financial documents.
The dueling releases, particularly Anthropic’s Cowork, kicked off an intense debate in the tech world about the rapid inroads of generative AI models, particularly Anthropic’s, into applications long dominated by software companies, with many heralding the “death of SaaS” at the hands of large language and other AI models.
Meanwhile, the staccato retaliatory strikes between Anthropic and OpenAI shows that the consumer and enterprise AI markets are, in some ways, intertwined.
“What’s happening in the enterprise is also what we call this consumerization of IT,” said Arun Chandrasekaran, an analyst at Gartner. “‘I am using this tool in my personal life. It is phenomenally helpful. And you know what? I want to use that in my workplace.’”
Because of that, even though Anthropic has found success in the enterprise market, it can’t ignore the consumer market, Chandrasekaran added.
A Push for Enterprise
On the other hand, while OpenAI is a leader in AI technology, the vendor appears unable to generate sufficient revenue, at least in the short term, without success in selling to enterprise customers. In the past few months, the vendor has partnered with SaaS providers such as ServiceNow and cloud data platform vendor Snowflake, indicating it is positioning itself in the enterprise market. Anthropic also rolled out a partnership with ServiceNow, the big IT self-service vendor.
OpenAI’’ release of platforms like Frontier is also an attempt to prove that it is enterprise-ready.
“It is a stake in the ground to say that, ‘Hey, we want to be hyper focused on the enterprise,'” Chandrasekaran said, adding that the vendor has shifted away from concentrating on future technologies such as superintelligence to more near-term projects. “They definitely want to focus more on the enterprise because they believe that both Google and Anthropic have had a really surprising surge in 2025 in terms of enterprise traction.”
However, proving to be enterprise-ready is not as easy as just releasing a product, Wang said. He said that while OpenAI is working to prove itself by hiring key executive talent such as Salesforce’s former Slack CEO, Denise Dresser, the model maker still has a long way to go.
“Culturally, they’re not ready,” Wang said. He added that for a vendor to be enterprise-ready, it must have clear, consistent roadmaps that enterprises can feel comfortable with. It must have customer service and on-site engineers who can help buyers succeed.
“You’ve got to make sure that these enterprise motions are consistent because a CIO’s and a CEO’s job is on the line, counting on your technology,” Wang said.
Compared with OpenAI, Anthropic and its chief commercial officer, Paul Smith, already have a plan in place to help enterprises succeed, Wang added.
“OpenAI, on the other hand, is just starting,” he said. “They’ll get there, but they don’t look like they’re enterprise-ready yet.”
Only an Ad
On the other hand, Anthropic won’t win this competition with a Super Bowl commercial.
“The really important moves that will happen from an OpenAI’s and Anthropic’s perspective will not be happening on television screens during the Super Bowl,” Nicholson said. “They’ll be happening in Zoom meetings and face-to-face meetings between the sales organizations of those two companies and the partners through which their capabilities will be delivered to enterprise customers.”
He added that it’s unlikely a CIO or CTO would be swayed by a Super Bowl commercial, even with more than 100 million pairs of eyeballs trained on it.
“What they will be swayed by is how effectively their existing SaaS providers are able to leverage these capabilities,” Nicholson added. “No one cares if it’s OpenAI or Anthropic, if what’s being delivered is an intelligent, authentic capability within my business and reasoning over my private data.”
Going Public and a Bigger Race
Nevertheless, Anthropic’s Super Bowl strategy will likely help shape public perception as the vendor races toward an IPO. The same holds for OpenAI, with its own ad and IPO aspirations.
“Whoever goes public first sets the tone in terms of the valuations, so they are all rushing to get out before Q4,” Wang said.
However, as both vendors fight it out in the enterprise and consumer arenas, a larger competition is playing out.
“Both Anthropic and OpenAI are now competing with Google Gemini for the enterprises,” Wang said.
And that high-stakes competition goes beyond Google, Chandrasekaran noted. The two feisty independents are also vying with hyperscalers such as AWS and Microsoft, as well as SaaS companies such as Salesforce, ServiceNow and Workday, and platform vendors such as Databricks and Snowflake, for enterprise attention in agentic AI and as the top enterprise choice.
“This is in some sense an expanded battle,” Chandrasekaran said. “This is really wanting to be that layer for building continuous workflows … in the future. All these companies are trying to put a stake in the ground to say that they are the platform that customers should look at.”
