'Built on Fear': OpenAI Slams Anthropic's Safety Culture in Private Memo

A leaked memo from OpenAI executive Denise Dresser has revealed a hostile stance toward rival Anthropic. The document accuses Anthropic of inflating its $30 billion revenue run rate by $8 billion through questionable accounting. Additionally, OpenAI slams its competitor's "fear-based" culture and lack of compute power.

A leaked internal memo from OpenAI's Chief Revenue Officer, Denise Dresser, has revealed the company's aggressive plan to take over the enterprise AI market. The document, first reported by The Verge, reveals that the company is no longer content with just leading the race. It shows OpenAI taking direct aim at one of its biggest rivals, Anthropic, accusing the firm of everything from creative accounting to being built on a culture of "fear" in the AI segment.

Accusations of Anthropic's "juiced" revenue in OpenAI leaked memo

One of the most interesting parts of the memo involves Anthropic's wallet. Reports from Bloomberg recently suggested that Anthropic's annualized revenue was trending over $30 billion. However, it seems OpenAI is throwing cold water on those figures. Dresser claimed in the memo that Anthropic uses "accounting treatment that makes revenue look bigger than it is."

According to the leak, OpenAI believes Anthropic is overstating its financial situation by roughly $8 billion. The memo alleges that Anthropic "grosses up" its revenue-sharing agreements with giants like Google and Amazon, rather than using net revenue figures. For Dresser, this is a tactical attempt by Anthropic to look more successful than it actually is, as both companies reportedly eye initial public offerings.

Anthropic "built on fear"

Dresser didn't stop at the balance sheet. She launched a philosophical attack, stating that Anthropic's entire story is "built on fear, restriction, and the idea that a small group of elites should control AI." This is a sharp jab at Anthropic's "safety-first" branding, which OpenAI now frames as a liability.

The memo further argues that Anthropic made a "strategic misstep" by failing to secure enough compute infrastructure. Dresser noted that customers are already feeling the consequences through service throttling and lower reliability. By contrast, she claims OpenAI acted faster on the "compute curve," giving them a structural advantage that Anthropic simply cannot match.

Tactical errors in a platform war

OpenAI also criticized Anthropic's heavy focus on coding assistants. While Dresser acknowledged that this gave them an "early wedge" in the market, she warned that "you do not want to be a single-product company in a platform war."

The memo outlines a shift in OpenAI's own strategy to capitalize on these perceived weaknesses. OpenAI is working to displace its rival in environments where Anthropic used to feel safe with moves like expanding reach via Amazon AWS Bedrock.