Anthropic Downgrades its AI Safety Policy Amid Market Pressures

    The AI vendor previously committed to releasing only models it classified as safe.