Yes. I'm happy to explain that a bit further.
As Professor Owen also mentioned, chatbots draw on many, many sources of information. One of them is journalistic and news organizations. In fact, companies such as OpenAI have signed deals with news organizations. Sometimes, when information is reproduced for users who ask a question, sources will be provided so that the user knows this came from The Wall Street Journal, The New York Times or whatever, but often that's not true.
What we've seen, of course, is that foreign actors seeking to interfere have evolved their tactics. One new tactic that was documented by the Institute for Strategic Dialogue is for Russian state media sources that have been banned in the European Union to produce massive volumes of information. It's super easy to do with generative AI. They produce a lot, and they look for spaces in which there isn't a lot of information already. This means that when chatbot providers are trawling through the Internet for information, they're picking up the only stuff there is on certain topics—which is Russian state media—and then reproducing it. It's basically a back door—