Artificial intelligence companies are growing "less transparent" as their new models grow "more powerful," Axios reported on Wednesday, stressing that AI is the "fastest-growing product category" in world history.
AI firm Anthropic is the fastest-growing company in the history of American business, with its revenue surging from $1 billion in late 2024 to $9 billion a year later, and $30 billion today, the report said.
It has also developed a highly powerful model, Claude Mythos Preview, capable of uncovering serious security flaws. Instead of releasing it publicly, Anthropic is limiting access to a small group of organizations to help defenders prepare, the report noted.
AI is also increasingly building itself. At Anthropic, nearly %100 of the code is AI-generated, according to the report. Fully autonomous AI research may arrive as soon as 2028, with recursive self-improvement seen as a major future risk.
At the same time, the report warned, the transparency is declining. Stanford's 2026 AI Index found that the most advanced models are now the least transparent, with little disclosure around how they are trained.
