And as soon as a correlation, not necessarily causation, to artificial intelligence is found, we implement liability frameworks. So for example in Taiwan right now, for some months now, if a deepfake scam asks people to invest in crypto or whatever is posted on a social media that earns money from advertisement — say Facebook — and Facebook even after they know that such thing exists, did not de-platform it, then after 24 hours Facebook is not fined, but if somebody gets conned 1 million dollars, Facebook in Taiwan is now liable for that 1 million dollars.