WASHINGTON (Sputnik) - US AI developer Anthropic has warned that malicious actors could potentially use powerful AI models to create chemical, biological, or other dangerous weapons, according to the company's registration statement for an initial public offering (IPO).