Anthropic Revenue Surges Past $30 Billion Annually, Valuation Discussions Reach $800 Billion to $1 Trillion

Anthropic has achieved annualized revenue exceeding $30 billion, potentially surpassing competitor OpenAI, while investors discuss valuations ranging from $380 billion to $1 trillion. The company has restricted access to its Claude Mythos cybersecurity model, claiming it possesses unmatched vulnerability detection capabilities. However, two new studies challenge this exclusivity, demonstrating that small, openly available AI models can reproduce most of the cybersecurity analyses Anthropic showcased for Claude Mythos. The research findings raise questions about whether Anthropic's restricted access policy is justified by truly unique capabilities or represents an overstated competitive advantage.

Claude Mythos Cybersecurity Capabilities Challenged

• ▪Anthropic claims Claude Mythos has capabilities that no rival can match.
• ▪Anthropic has kept its Claude Mythos cybersecurity model on a short leash.
• ▪Two new studies suggest that small, openly available models can reproduce most of the vulnerability analyses Anthropic has showcased for Claude Mythos.

Open-Source Models Replicating Vulnerability Detection

• ▪Two new studies examined the cybersecurity bug detection capabilities of small open models compared to Claude Mythos.

Perspective of Anthropic

• ▪Anthropic has restricted access to Claude Mythos due to concerns about its cybersecurity capabilities.
• ▪Anthropic maintains that Claude Mythos possesses cybersecurity capabilities unmatched by competing AI models.

Perspective of AI safety researchers and open model advocates

• ▪Research findings suggest Anthropic's restricted access to Claude Mythos may not be justified by unique capabilities.
• ▪Small open-source AI models can perform vulnerability detection tasks comparable to Claude Mythos according to recent studies.

2 sources