how do you decide what is and isn't allowed? it's a free speech world, but is that harmful? so we worked with 150 expert agencies. europol, the home office, institute for strategic dialogue and so on, to come up with some policies to define that. we then trained people to classify videos, and then when they could do it accurately, we trained machines and now we report on how effective that policy is. and i can tell you that 90% of videos that violate that policy never are seen by a single human. so we can do that in the context of a legal framework, but then codes of practice and transparency with regulators. and i think that's the way forward with 0nline safety bill. set some clear responsibilities, but then you need to develop policies which will change over time. so would you say the system is working at the moment in terms of protecting people? no, because i think that's, we've done that ourselves, but that's not consistently a standard across the different tech platforms. and we see, you know,