On April 20, 2026, OpenAI Chief Revenue Officer Denise Dresser distributed an internal memorandum to employees that directly challenged the business practices and core philosophy of primary competitor Anthropic. The memo, which circulated throughout the organization’s internal communications channels, marks a significant escalation in the private rivalry between the two leading artificial intelligence firms. Dresser’s critique focused on three primary areas: financial transparency, computational efficiency, and the ideological framework governing AI development.
In the memorandum, Dresser specifically questioned Anthropic’s revenue accounting methods. While Anthropic has recently reported significant growth in its enterprise sector, Dresser alleged that the company’s figures may be bolstered by non-standard accounting practices or one-time credits that do not reflect sustainable recurring revenue. She suggested that OpenAI’s own revenue streams, which have scaled significantly following the release of its latest enterprise-grade models, are built on a more robust foundation of direct consumer and corporate subscriptions. Dresser did not provide specific internal financial data for Anthropic but indicated that OpenAI’s internal analysis suggested a discrepancy between reported growth and actual market penetration.
The memo also addressed the technical and operational strategies employed by Anthropic. Dresser claimed that Anthropic’s compute strategy is less efficient than OpenAI’s, asserting that the rival firm requires a disproportionate amount of processing power to achieve performance benchmarks comparable to OpenAI’s flagship models. She noted that OpenAI’s focus on architectural optimization has allowed it to maintain a competitive edge in cost-per-token metrics, a key figure for enterprise clients seeking to integrate large language models at scale. This focus on efficiency was presented as a core differentiator in OpenAI’s ability to manage its massive infrastructure costs.
Beyond financial and technical metrics, Dresser launched a sharp ideological critique of Anthropic’s corporate identity. She accused the company of leveraging a narrative of fear and restriction to differentiate itself in the market. According to the memo, Dresser stated that Anthropic’s approach to AI safety is rooted in the idea that a small group of elites should control AI, rather than the more open and democratized vision she attributed to OpenAI. This philosophical divide has been a point of contention since Anthropic was founded by former OpenAI executives who departed over disagreements regarding the company’s commercial direction and safety protocols.
Dresser’s memo concludes by urging OpenAI staff to remain focused on the company’s product roadmap and deployment goals. As of the date of the memo, OpenAI has not issued a formal public statement regarding the internal document, and Anthropic has not provided an official response to the specific allegations regarding its revenue accounting or compute efficiency. The internal communication highlights the intensifying competition for market share and talent within the generative AI sector as both companies seek to secure long-term dominance in the enterprise software landscape.