00:00The issues connected to AI that MI5 with our various partners is dealing with are about the potential for AI
00:07to be misused by those who already mean the UK harm.
00:10So that would include terrorists trying to use AI to improve their reconnaissance methods, for example.
00:17It would include serious criminals, not MI5's own issue but for our colleagues in the National Crime Agency.
00:22It would include attempts to use AI to create bespoke radicalising material sort of tailored to the individual.
00:30So there's a series of issues where those who are already trying to present threats to the UK might seek
00:35to use AI to make those threats a bit sharper.
00:38And our job, therefore, is to stay ahead of those adversaries and make sure that we are using AI to
00:43enhance our security rather than allow it to be undermined.
00:46In the particular case of bioweapons, I was very clear in my speech today that this is a potential risk
00:51rather than something that we are already seeing manifesting.
00:54So it needs sober work now to prepare for the possibility that either AI could enable dangerous knowledge, for example,
01:03around bioweapons to reach people who mean us harm in a way that they found hard to access until now.
01:09And or there is also potential for AI to be used to generate new pathogens in the first place.
01:15podemos eat better手 for free and c run the nggak.
01:15We'll eat better Verniz.
01:15We'll eat better methods here.
01:15We will be better on the next slide.
01:16We'll be better conversations.
01:16You
