The hell is going on with this article, is this bot-written? The top-line reads that the CCDH are the ones running the analysis. But the very next line reads "Streaming Platform YouTube said they analysed over 12,000 videos across 96 channels using an AI model crafted specifically to be able to distinguish between reasonable scepticism and false information." So it kinda sounds like this should be titled "YouTube study investigates changes in climate denial rhetoric, finds deniers are succeeding at skirting older protections." and then go on to explain that the new model inherently identifies this problematic content.
Listen, I'm not a big fan of Google, but as written this is just a shitty hit piece arguing in favor of an activist group that seems to be calling on YouTube to do the thing they've just said they already did. Unless the claim is that YouTube just went "Huh, weird. Guess we'll keep making money on it anyways!" and there's proof of that, this feels pretty deliberately misleading.