Anthropic has investigated AI consciousness and model welfare, with CEO Dario Amodei stating in February 2026 that the company does not know if its models are conscious