Battle of the AI brands: What’s behind the bad blood between OpenAI and Anthropic

As more than 100 million people watched the Super Bowl, the battle for the future of artificial intelligence spanned a huge public space.

In a series of viral ads, Anthropic, a newcomer to the AI ​​field, broadcast to soccer fans that they should avoid AI with ads. The ad was obviously aimed at OpenAI, which plans to add ads to ChatGPT. Anthropic’s robot, Claude, will never have advertising, he says.

In one of the Super Bowl slots, a short man asks an AI assistant how to get six-pack abs. The very supportive AI trainer says he has a training plan for the man before abruptly pivoting to pitch for height-increasing insoles.

Sam Altman, co-founder and chief executive of OpenAI, hit back on social media, calling the ads “clearly dishonest.”

The public feud between the dueling San Francisco companies has highlighted an intense rivalry between OpenAI and Anthropic as they face off in a battle that could shape the future of technology. Although both companies are still very young, their fight could become the next big clash between brands like IBM vs. Apple, Coca-Cola vs. Pepsi or McDonald’s vs. Burger King.

While there are other AI giants, including Google, Meta, Microsoft, Amazon, and others, the tension between OpenAI and Anthropic is telling because it stems from their shared origins and different philosophies.

OpenAI has always favored fast, free public releases of its latest AI models, such as ChatGPT. It believes that feedback from many users helps to improve its models and make them safer.

Anthropic, co-founded by Dario Amodei, who left OpenAI in part because of its focus, says it is taking a more cautious approach. He says he’s focused on how to build AI products slowly and safely. It tries to control who gets the latest technology before releasing it to the general public.

This difference in position – OpenAI’s need for speed versus Anthropic’s emphasis on security – helped spark the tension. Experts say OpenAI seems more likely to push the boundaries with broad public access, while Anthropic initially prefers incremental steps through partnerships. Everyone wants to steer the conversation around what responsible AI will mean.

Of course, it’s also a fight for market share in a rapidly growing, multi-billion dollar industry. The two companies compete for business and individual customers. They both want to become thought leaders and revenue leaders in the AI ​​space.

“AI is the most important technology we have ever known in human history,” said Tim Law, research director for AI and automation at IDC, a global business intelligence company. “They compete on so many different levels.”

Tension between Altman and Amodei appeared to surface last week, when they were the only ones not to shake hands on stage during a group pose at an AI event in India.

Amodei and Altman also have different views on AI’s potential to displace workers. Amodei said by 2025, AI could eliminate half of all entry-level white-collar jobs. Altman acknowledged technology’s role in disrupting jobs, but says it can also create new, more fulfilling roles.

Amodei previously worked at OpenAI as vice president of research. He and other OpenAI employees, including his sister Daniela, left the company to create Anthropic in 2021 due to disagreements over OpenAI’s direction.

In a 2024 interview, Amodei said he had a different vision for how AI safety should be handled, and “it’s incredibly unproductive to try to challenge someone else’s vision”, so he left.

Both Anthropic and OpenAI offer chatbots to consumers. Anthropic, however, has committed that Claude will remain ad-free while OpenAI is currently testing ads in ChatGPT.

Valued at approximately $380 billion, Anthropic is legally required to balance making a profit with promoting the company’s public benefit from the “responsible development and maintenance of advanced AI for the long-term benefit of humanity.”

It makes money through corporate contracts with companies, government agencies and other organizations, as well as paid subscriptions. Its chatbot, Claude, has fewer users than ChatGPT, but developers and businesses, including Amazon, Microsoft and other companies, often prefer Anthropic’s tools. Claude can generate code, translate languages, write and perform other tasks, but the chatbot cannot produce images.

It claims to have revenues equivalent to around $14 billion per year. PitchBook estimates it has about 2,000 employees.

Anthropic pushes for AI safety through political donations. This month, the company announced it would contribute $20 million to a group called Public First Action to “support public education about AI, promote protective measures, and ensure America leads the AI ​​race.” The organization advocates for greater regulation of AI’s high risks, such as its potential use in biological weapons and cyberattacks.

Anthropic faces pressure to reduce its guarantees. The Pentagon reportedly threatened to terminate its $200 million contract with Anthropic if the company did not ease restrictions on its AI model so it could be used for more military purposes.

The company this week abandoned its pledge not to release AI models if Anthropic cannot guarantee it could properly mitigate risks amid increased competition. It is unclear whether the change is related to the Pentagon contract.

OpenAI’s valuation exceeds $800 billion. It has about 4,000 employees, according to PitchBook. The company’s chatbot can also code, write and generate images. More than 800 million people use ChatGPT every week for writing, brainstorming, coding, and other tasks.

Originally founded in 2015 as a nonprofit, the company restructured to raise more capital. While testing ads in ChatGPT, OpenAI also makes money in other ways, including fees and subscription plans.

Although the companies overlap in terms of customers, OpenAI has a much larger user base that also uses ChatGPT for free, making advertising an attractive revenue stream.

“The reality is that advertising is central to the entire Internet business model,” said Andy Wu, an associate professor at Harvard Business School. “Given OpenAI’s consumer focus, I think it’s inevitable that OpenAI will have to advertise.”

OpenAI has started testing ads in ChatGPT and says conversations will remain private to advertisers.

OpenAI’s Altman said Anthropic’s Super Bowl ads illustrate how elitist it is.

“Anthropic is offering an expensive product to the rich,” Altman posted on

Despite the jabs thrown by both companies, industry insiders say it’s healthy for these debates to spill over into the public, even if it could add to confusion about the power of the technology and the big companies behind it.

“This is going to have significant consequences for society,” IDC Law said. “This could be a very disruptive time for businesses, as well as consumers and governments.”