I’ve seen a number of fake chat reports with doctored screenshots. It tends to be quite easy to spot when you know what you’re looking for. Even when done well, more often than not the phrase being reported isn’t able to get past the chat filter anyway. Those that do submit faked evidence are barred from reporting in future. There are a number of filters in a test state at any one time, we aren’t actively blocking but we are testing their accuracy, just because a faked sentence gets through the filter doesn’t mean it isn’t logged.