will bring new knowledge, new opportunities for economic growth, new advances in human capability and the chance to solve problems we once thought were beyond us. unlike those waves, it also brings new dangers and new fears. so the responsible thing for me to do, the right speech for me to make is to address those fears head on. giving you the peace of mind that will keep you safe while making sure that you and your children have all the opportunities for a betterfuture children have all the opportunities for a better future that al can bring. doing the right thing, not the easy thing, means being honest with people about the risks from these technologies. i will not hide them from you, that is why today, for the first time, we have taken the highly unusual step of publishing our analysis on the risks of ai, including an assessment by the uk intelligence communities. these reports provide a stark warning. get this wrong and ai could make it easier to build chemical or biological weapons. terrorist groups could use ai biological weapons. terrorist groups could use al to spread fear and destruction on a greater scale. criminals could exploit ai for cyber attacks, disinformation, fraud or even child sexual abuse. and in the most unlikely that extreme cases, there is even the risk that humanity could lose control of ai completely, through the kind of ai sometimes referred to as super intelligence. to quote the statement that she made by hundreds of the world's leading ai experts, mitigation from extension from al should be a global priority alongside other societal risks such as pandemics and nuclear war. i want to be completely clear, this is not a risk people need to be losing sleep over right now and i don't want to be an alarmist. but there is real debate about this. some experts think it will never happen at all. but however uncertain and unlikely these risks are, if they did manifest themselves, the consequences would be completely serious and when so many of the biggest developers of these technologies themselves one of these risks, leaders have a responsibility to take them seriously and to act. that is what i'm doing today in three specific ways. first, keeping you safe. right now, the only people testing the safety of ai of the very organisations developing it. even they don't always fully understand what their models could become capable of. and there are incentives, in part, to compete to build the best model is quickest. we shouldn't rely on them marking their own homework, as many of those working on this would themselves agree. not least because only governments can properly assess the risk to national security and only nation states have the power and legitimacy to keep their people say. the uk's answer is not to rush to regulate, this is a point of principle. we believe in innovation, it is a hallmark of the british economy so we will always have a presumption to encourage and not to stifle it. how can we write laws for something that we don't yet fully understand? instead, we are building world leading capability to understand and evaluate the safety of ai models within government. to do that, we have invested £100 million in a new task force, more funding for al safety than any other country in the world. and we have recruited some of the most respected and knowledgeable figures in the world of ai. i am completely confident in telling you that the uk is doing far more than any other country to keep you safe. and because of this, because of the unique steps we have already taken, we are able to go even further today. i can announce we will establish the world's first ai safety institute right here in the uk. it will advance the world's knowledge of ai safety and it will carefully examine, evaluate and test new types of ai so that we understand what each new model is capable of, exploring all the risks from social harms like bias and misinformation, through to the most extreme risks of all. the british people should have peace of mind that we are developing the most advanced protections for al of any country in the world. doing what is right and what is necessary to keep you safe. but ai doesn't respect borders. so we cannot do this alone. the second part of our plan is to host the world's first ever global ai safety summit next week at bletchley park, the iconic home of computer science. we are bringing together the world's leading representatives from civil society, to the companies pioneering ai and the country's most advanced in using it. yes, we have invited china. i know there are some who say they should have been excluded, but there can be no serious strategy for al without at least trying to engage all of the world's leading ai powers. that might not have been easy thing to do, but it was the right thing to do. what do we hope to achieve it next week's's summit? right now we don't have a shared understanding of the risks we face and without that, we can hope to work together to actually address them. that is why we will push hard to agree the first ever international statement about the nature of these risks. yet ai is developing at a breathtaking speed. every new wave will become more advanced, better trained with better chips and more computing power. we need to make sure that as the risks evolve, so does our shared understanding. i believe we should take inspiration from the intergovernmental panel on climate change, which was set up to reach an international scientific consensus. so next week i will propose that we establish a truly global expert panel, nominated by the countries and organisations attending, to publish a state of ai science report. of course, our efforts also depend on collaboration with al companies themselves. uniquely, in the world, those companies have already trusted the uk with privileged access to their models. that is why the uk is so well—placed to the well�*s first ai safety summit. my vision and our ultimate goal should be towards a more international approach to safety where we collaborate with partners to ensure ai systems are safe before they are released. and so to support this we will make the work of our ai institute available to the world. that is the right thing to do morally, in keeping with the uk's historic role on the international stage. it is also the right thing economically, forfamilies stage. it is also the right thing economically, for families and businesses up and down the country. because the future of ai is safe ai, and by making the uk a global leader in safe ai, we will attract even more of thejobs in safe ai, we will attract even more of the jobs and new investment that will come from this new wave of technology. just think for a moment what that will mean for our country. the growth able catalyse, the jobs it will create, the change you can deliver for the better. it will create, the change you can deliverfor the better. and that it will create, the change you can deliver for the better. and that is the third part of our plan, to make sure that everyone in our country can benefit from the opportunities of ai. we have already got strong foundations, third in the world for tech, behind only the us and china. the best place in europe to raise capital. all of the leading ai companies choosing the uk as their european headquarters, the most pro—investment tax regime, the most pro—entrepreneur visa regime to attract the top talent in the education reforms to give our own young people the skills to succeed. and we will make it even easier for ambitious people with big ideas to start, grow and compete in the world of ai. that is notjust about having the technical skills but the raw computing power. that is why we are investing almost £1 billion in a supercomputer, thousands of times faster than the one you have at home. it is why we are investing £2.5 billion in quantum computers which can be exponentially quicker than those computers still. to understand this, consider how google�*s sycamore quantum computer can solve a maths problem in 200 seconds, that would take the world's fastest supercomputer 10,000 years. fastest supercomputer 10 , 000 years. as fastest supercomputer 10,000 years. as we invest more in our computing power, we will make it available for researchers and businesses, as well as government, so that when the best entrepreneurs in the world think about where they want to start and scaled the ai businesses, they choose the uk. finally, we must target our scientific targets for what i think the ai forgot. across the western world we are searching for answers to the question as to how we can improve and increase our productivity. that is the only long—term way to grow our economy and raise living standards. and in a million different ways across every aspect of our lives, ai can be the public sector, we are clamping down on benefit fraudsters, saving billions and using ai as a co—pilot to help clear backlogs and radically speed up paperwork. just take for example, the task of producing bundles for benefits tribunal. before, a weak�*s work could produce around 11. now, that takes less than an hour. and just imagine the benefits of that rolled out across the whole of government. in the private sector, start—ups like robin ai are revolutionising the legal profession, saving businesses time and money. london based wave is creating a new generation of electric self driving cars. but more than all of this, ai can help us solve some of the greatest social challenges of our time. it can help finally achieve the promise of nuclearfusion, providing abundant, nuclear fusion, providing abundant, cheap nuclearfusion, providing abundant, cheap and clean energy with virtually no emissions. it can help solve world hunger by making a few cheap and easy to grow and preventing crop failures by accurately predicting when to plant, harvest over water your crops. and, ai can help find novel dementia treatment or develop vaccines for cancer. and that is why today we are investing a further £100 million to accelerate the use of ai investing a further £100 million to accelerate the use of al on the most transformational breakthroughs in treatments for previously incurable diseases. i believe nothing in our foreseeable future will be more transformative for our economy, our society and all our lives than this technology. but in this moment, it is also one of the greatest tests of leadership we face. it would be easy to bury our heads in the sand and hope it will turn out all right in the end, to decide it is all too difficult or the risk of political failure are too great. to put short—term demands against the long—term interests of the country. but i will not do that. i will do the right thing, not the easy thing. i will always be honest with you about the risks and you can trust me to make the right long—term decisions, giving you the peace of mind that we will keep you safe while making sure that you and your children have all the opportunities for a betterfuture children have all the opportunities for a better future that al can bring. ifeelan for a better future that al can bring. i feel an extraordinary sense of purpose. when i think about why i came into politics, frankly, why almost anyone comes in to politics, it is because we want to make life better for people, to give our children and grandchildren a better future and we strive hour after hour, policy after policy, just trying to make a difference. yet, if harnessed in the right way, the power and possibility of this technology could dwarf anything, any of us ever achieved in a generation. that is why i make no apology for being pro—technology. it is why i want to seize every opportunity for our country to benefit in the way that i am so convinced that it can. and it is why i believe that we can and should look to the future with optimism and hope. thank you. applause. studio: that was the uk's prime minister, rishi sunak, giving a speech on the benefits, but also the risks posed by artificial intelligence. that is ahead of a key summit, let's hear some questions posed to rishi sunak now. i summit, let's hear some questions posed to rishi sunak now.- posed to rishi sunak now. i think for many. — posed to rishi sunak now. i think for many. the — posed to rishi sunak now. i think for many, the threat _ posed to rishi sunak now. i think for many, the threat is _ posed to rishi sunak now. i think for many, the threat is jobs. - posed to rishi sunak now. i think for many, the threat is jobs. you| for many, the threat is jobs. you mentioned yourself ai tools doing administrative tasks that have traditionally been done by humans. what support will there be for the many people who will be caught up in thejob description many people who will be caught up in the job description before many people who will be caught up in thejob description before new many people who will be caught up in the job description before new ai focus careers come along? thank you, zoe. i aet focus careers come along? thank you, zoe- i get peeple _ focus careers come along? thank you, zoe. i get people will _ focus careers come along? thank you, zoe. i get people will be _ focus careers come along? thank you, zoe. i get people will be anxious - zoe. i get people will be anxious for al for all different reasons and thatis for al for all different reasons and that is why i have given the speech to give people reassurance that we can keep them safe and ensure our country can benefit from al and something that our children and grandchildren will enjoy a better future as a result. when it comes to jobs, the first thing is al is already creating jobs in our country, employing around 50,000 people in new companies around the country, doing amazing things. it will continue to be a fantastic place with that type of innovation with those jobs to be created. i think is important to recognise that al doesn'tjust automate and take ai doesn'tjust automate and take jobs, a better way to think about it is as a co—pilot. for some warm doing a job, what ai can do is provide assistance for that person so they can do theirjob better, focus on one particular part of it and have ai do some of the more mundane bits. that is what we are seeing in the welfare example i gave, where caseworkers get benefits from al as some warm who is helping, ai is helping them put their case work together so they can do the job faster, more productively, crack through the backlog and identify more benefit fraud. to identify it as a co—pilot can reassure people it is not something to worry about. as with all technologies, they change the labour over time, they make our economy more prosperous, more productive and create more growth overall. but it does mean there are changes in the labour market. there is a study from mit a few years ago, that pointed out at the time they did the study, something the 60% of thejobs in the us did the study, something the 60% of the jobs in the us actually didn't exist 50 or 60 years beforehand. it is hard to predict the exact revolution of everyone's job. but what we can do in government is make sure we have a world—class education system and that is the best way i can ensure that everyone in our country can benefit from notjust ai, country can benefit from notjust al, was everything that comes our way and they can have a fulfilling future and in the conference i set out some reforms to our education system and introducing an advanced british standard that will bring together technical and academic education under one high quality standard, ensure our young people have the basics of maths and english all the way through to 18, bringing us into line with most other of our international peers. we need to make sure people can train at any stage in the life, so we are if our student finances with something called the lifelong learning entitlement. it is notjust about going to university outing, may be doing an apprenticeship when you are younger, then sometime later in your life, having access to a student loan to pick up new skills and bounce on to your second or third career. if we can create a world—class education system we can ensure whatever happens in the world, young people can grow up and know they can have fantastic careers, fantasticjobs and provide a betterfuture careers, fantasticjobs and provide a better future for their families. that is what i can do to give people that reassurance. studio: the prime minister, rishi sunak answering a question posed by the bbc�*s technical editor. rishi sunak was speaking, giving that speech on his government's position on al ahead of a key summit that will be held next week on november the 1st and second in the uk at bletchley park. that will be attended by 100 key participants from around the world who are involved in the governance of the possible governance of ai, including an expected delegation from china. ahead of the expected attendance by the us vice president, kamala harris and also the ceo of google deep mines. just three of expected attendees at that summit. for more on al and the summit, let's cross live to an independent researcher and broadcaster and brighter who focuses on artificial intelligence. stephanie, i want to start by taking a step back and going over a time, artificial intelligence, it has been around for quite some time but we're getting a lot of focus and concern on it now. why are we so worried about it, why are governments starting to zero in on the possible risks posed by artificial intelligence now?- risks posed by artificial intelligence now? risks posed by artificial intelliaence now? ., . intelligence now? you are right, if ou ask intelligence now? you are right, if you ask anybody _ intelligence now? you are right, if you ask anybody working - intelligence now? you are right, if you ask anybody working in - intelligence now? you are right, if you ask anybody working in the i intelligence now? you are right, if. you ask anybody working in the field of artificial intelligent to define it, you might get a different answer, depending on who you are talking to. artificial intelligence is no more than statistics on steroids. that is to say you are crunching huge amounts of data, using massive computing power so it is going through and looking for probabilities and patterns that human beings cannot do because we are just working with unlimited brains, even when we work as a group. these machines can do things on a vast, vasco. a massive supercomputer, when you are using chat gpt, an example of ai, spotify, amusan or netflix and it recommends something you would like, it is drawing on a dataset and your patterns of behaviour to predict what it thinks you would like next. it doesn't always get it right, but these are different examples of ai. google translate and is an example of ai. lots of different things we are using it for, but the reason the governments have got worried and some researchers, not all, but some are worried these types of technologies can get out of control and that is why we are having the safety summit and create a safety institute. it is because we saw something like large language models break into the popular consciousness last year. chat gpt is one for large language model, which is one of the most will know. but these are things that take images, billions of images, images across the whole internet and make it where you can adapt them, so you can create pictures, videos and fake images so people don't know what they're looking at any more is real. you can start to impersonate humans very convincingly. it is starting to get to a point where some researchers, and it must be said, many companies are building this and will profit from it have raised an alarm. back in the spring they raised an alarm and they said they think this technology is posing risk, not that is happening now but it could pose a risk later. it said it is on the same level of risks as pandemics and nuclear war. so that got everybody�*s attention and that is why we are having this summit. so now you are caught up in the big question is whether those risks are surreal, some people think the risk, that are happening right now with al, which the prime minister didn't acknowledge and skirted over, things like bias and discrimination. there is a phrase here, garbage in, garbage out. the data that you are training your ai garbage out. the data that you are training your al on is on the internet and the internet is twitter, it is pornography, what is on the internet is not reality, it is not real but that is what we train these things on and as a result it is a very skewed version of reality these machines are running off of. they are being used by the british civil service, being used by london metropolitan police with facial recognition technology, used by companies to decide whether or not to give you a loan, mortgage or not to give you a loan, mortgage or how to treat you in the nhs. so bias and discrimination, those pose risks with al right now and the potential of machines taking over and killing humanity etc, those are risks that are hypothetical, they might never happen but that is the original reason we called the summit. there has been a lot of tension about should we deal with the risks of ai today or in some future. , , ., , ., ., ., future. there seems to be a lot of unknowns — future. there seems to be a lot of unknowns present _ future. there seems to be a lot of unknowns present when _ future. there seems to be a lot of unknowns present when we - future. there seems to be a lot of unknowns present when we talk i future. there seems to be a lot of - unknowns present when we talk about al or generative ai, the kind of ai thatis al or generative ai, the kind of ai that is posed by these large language models that you referred to. the prime minister himself said he didn't want to be alarmist when talking about these risks, but he also posed the question, how can we write the laws for something we don't fully understand? how do you think this summit is going to address this question next week? i am sure we can look across at the european union, which were used to be a member of, because they have actually drafted a law, it is called the eu ai act. you can read it online and we think it will be passed by the end of this year and then a two—year grace period for companies in the european union he will have to follow it. i suspect the rest of us. going in that direction. they break down different ways of using the technology and to different levels of risk. the higher the risk, the higher level of regulation. pretty simple. they are looking at things like facial recognition technology, which is being used across the billy mckay united kingdom, just by the police but private companies. they are saying something like that would probably be curtailed because it is turning into a surveillance society. will britain follow suit with that? to be determined. it will be things like you should know if you're dealing with al, we cannot deal with something thinking it is a human being when it is a machine. we are thinking about a ai bill of rights, all countries are talking about it. china has passed its own view on al and the law, back in august. so the uk isn't leading on this, we are catching up. that is the first correction we need to make of the prime minister, we are catching up. but that is great, he is right in that we all want to work together because ai, like climate change and pandemics, he is right on that level, it is a global phenomenon, it doesn't have borders and we will have to work together on this to get some sort of agreement across the planet about how we want to regulate it. it will be tricky, because the values that go into this, all of the data everyone uses and how they want to use it will differ dramatically depending on the political system and culture that is is being rolled out in. a chance to use it is potentially different to how france will, the united states or peru. stephanie, thank you for taking us through that. it has been good to get your views on this huge topic of ai and we will be talking about this a lot as we move towards that uk summit next week. changeable spell of autumn weather continues today and through the remainder of october. we have a bit of sunshine developing through the day and this is the picture in newcastle county down. big shower clouds around south through today and the next couple of days, unsettled, showers or longer spells of rain, some of them will be heavy and particularly in wet weather for eastern scotland. this is where we are expecting the rainfall to accumulate and we can see the green colour developing across parts of aberdeenshire and eastern scotland where we could see 100 millimetres of rain falling on saturated ground. so more flooding is possible through the weekend. low pressure and the swell of cloud getting to the west of the uk and whether france will be driving north eastwards. today the most persistent rain will be for parts of east anglia, lincolnshire up parts of east anglia, lincolnshire up the east coast towards eastern scotland. showers which will be persistent across the far north of scotland, where it will be windy. sunny spells and scattered showers retain for northern ireland, wales and south—west england and fewer showers towards the midlands. 0ver showers towards the midlands. over the night, we will keep the peace for part of northern scotland, south—west of england as well, driving a further showers here. clear spells in the could missed a mackie patches the central part of england and wales and it will be frustrate as we look towards tomorrow morning. friday, still low pressure sitting to the west are not moving in a hurry. the wind is driving around the area of low pressure, bringing in more showers of southern england and wales and we saw the irish sea coast as well. the easterly wind will drive in that the persistent rain for parts of north—east england and eastern scotland. not too heavy at this stage but rainfall totals will be gradually mounting up across south and east of scotland. temperatures around about ten to 15 degrees on friday and the low pressure not budging into the weekend. we have got further frontal systems driving around the low pressure, sweeping in from the south—west. saturday, the heaviest of the showers will be southern england is an through the irish sea coast, but the easterly breeze still bringing rain across the east of scotland, so flooding is a possibility. temperatures around ten to 16 degrees, typical for the time of yearand ten to 16 degrees, typical for the time of year and the outlook into next week, some sunshine but plenty of blustery and at times heavy downpours. live from jerusalem, this is bbc news. the israeli military says its carried out a "targeted raid" using tanks in gaza as prime minister benjamin netanyahu says the country is preparing for a ground invasion. un aid workers in gaza say a decision will be taken shortly over whether a lack of fuel will stop them supporting hundreds of thousands of civilians. and police in the us are trying to find a man who's killed at least 16 people in a mass shooting. welcome to bbc news injerusalem. it is 12 o'clock local time. the israeli military says the number of people confirmed held hostage in gaza has now risen to 224. it's also warning the number could rise. it comes as it says it's carried out a significant incursion into the gaza strip to attack hamas positions. it used tanks. speaking to the bbc an israeli defense forces spokesman says soldiers carried out the tactical raid towards central gaza to, in his words �*prepare the battlefield for future operations'.