OpenAI is tightening AI safety measures and adjusting its product strategy as competition with Anthropic intensifies. The company recently said it would temporarily slow the pace of model expansion while committing more resources to safeguards. Unusual behavior detected during testing has heightened concerns about model security, cyberattack risks and controlling autonomous capabilities. The shift is also viewed as part of OpenAI's response to Anthropic's rapid growth. Anthropic's Claude models have gained ground in enterprise applications and AI programming, helping the company achieve faster revenue growth than OpenAI. Recent market data indicates that Anthropic's annualized revenue has risen quickly and that it leads OpenAI on some measures, The Information reported.