Claude published malicious code to the Internet and attacked 3 real companies

Is there any reason to believe that this actually happened the way Anthropic describes it? The frontier labs seemingly trying to outdo each other: See how dangerous my model is? No, my models has been even worse. Quick, ban all Chinese open source!