Day after Anthropogenic filed confidentially to be made public, moving forward OpenAI IN a closely watched IPO race among AI giants, Daniela Amodeithe company’s co-founder and president, tried to draw a sharper distinction between the two rivals. Anthropic is currently valued at $965 billion, with expectations that it could exceed $1 trillion in public markets, compared to OpenAI’s valuation of around $900 billion. But as Amodei described it, competition isn’t just about numbers, it’s about how the technology itself is built and used.
“The whole reason we started Anthropic is to be able to build and develop this technology in a way that’s ethical, responsible, fair, and I think that’s really the job of everyone in the company, but especially the leadership, to say, all these numbers, they’re not actually dots,” Amodei said at this year’s Bloomberg Technology Summit yesterday (JJ in San Francisco 4).
Anthropic was founded by seven former OpenAI employees, including Daniela and her brother, Dario, who is the company’s CEO, aiming to build a more transparent firm focused on AI security. Its divergence from OpenAI extends beyond positioning to how it plans to grow.
The company has emphasized providing computing capacity, including a deal with Elon Musk’s SpaceX (which is absorbed xAI earlier this year) to access its data centers in Memphis that will cost $1.25 billion per month. But Amodei said Anthropic is deliberately avoiding the aggressive spending levels seen elsewhere. OpenAI has predicted up to $600 billion in computational spending by 2030; Anthropic expects to spend roughly a third of that.
“The structure of these deals is that you have to commit to a certain amount of computing well in advance, and (we don’t want to overdo it) so that we buy more computers than we can productively use,” Amodei explained. “We’d rather be on the side of slightly more demand for product than we can serve than the other way around, where you overshoot and then you’re not in a good situation because you bought something you can’t pay for down the road.”
While Anthropic has also expressed interest in more speculative infrastructure, as proposed by SpaceX orbital data centers, Amodei said there are “no immediate plans to work with astronauts to operate space data centers.” “But you never know,” she added.
Product strategy marks another key division. Anthropic has prioritized enterprise use cases and coding over mass market consumer engagement. This is in contrast to OpenAI, where more than 70 percent of ChatGPT usage it is related to personal tasks such as research, tutoring and life advice.
“We have always felt that entrepreneurship and business are the best spiritual fit for Anthropic and our values,” said Amodei. “The difference in our consumer product compared to competitors is that we’re not an entertainment tool. It’s really about productive activities, whether that’s at work or at home.”
Both companies, however, are investing heavily in advanced cybersecurity AI Claude Mythos of Anthropic has raised concerns about its ability to exploit vulnerabilities. OpenAI’s Daybreak targets similar risks, but takes a different approach to deployment.
Daybreak is integrated into existing GPT workflows and is offered in scalable access based on user verification. Mythos operates as a closed consortium limited to vetted organizations in approximately 15 countries, including the US government, NATO, ENISA, Samsung and Okta.
“You have to give defenders a head start,” Amodei said. “AI models will continue to advance. If it’s not us that one day releases a Mythos-level model (to the public), another AI company will.”
Anthropic has also taken a more cautious stance on government work. The company pulled out of a Pentagon contract involving domestic surveillance and autonomous weapons, which OpenAI later picked up. However, Amodei described the broader cooperation with the US government as positive.
“Each company will have its own principles of what its red lines and values are,” she said. “It’s important that, whatever those values are for you as a company, that you’re true to them, that you feel like you can explain them to your employees and to the world more broadly.
