Written questions by Spencer.

With reference to paragraph 2.40 of the Autumn Budget 2024, HC 295, published on 30 October 2024, if she will make an assessment of the potential impact of changes to employer National Insurance contributions on special educational needs transport providers.

Home-to-school travel is an integral part of the school system. It provides a valuable service on which many families rely. The department is grateful to the many transport operators for the crucial role they play in ensuring that children receive the education that they need to help them thrive.Departmental officials engage regularly with local authorities to understand the challenges they face and will continue to monitor this situation. The department does not expect there to be a significant impact on home-to-school travel for children with special educational needs and disabilities. Local authorities are responsible for arranging home-to-school travel and deliver this through a range of in-house services and external providers.The government recognises the need to protect the smallest employers. It has more than doubled the Employment Allowance to £10,500, which means that more than half of businesses with National Insurance liabilities will either gain or see no change next year. Businesses will still be able to claim employer National Insurance contributions relief, where eligible.

Food and Rural Affairs, pursuant to the Answer of 23 October 2024 on Bridges: Runnymede and Weybridge, if he will make an assessment of the potential impact of the Environment Agency's response to the closure of bridge 142 on public safety.

The Thames towpath provides health and wellbeing benefits to many communities, as well as being an important access route. The Environment Agency’s (EA’s) towpath bridge refurbishments are considered within their overall Navigation capital programme. Each project is evaluated against its impact on customer safety, income generation, water level management, design complexity, cost, and impact on the Navigation service. This programme is refreshed each year to maximise the benefits for Navigation customers within the grant-in-aid allocation the EA receives from the Government. The EA is now exploring how partnership funding could be used alongside the grant-in-aid funding it receives from the Government for bridge refurbishment projects. In the meantime, temporary security fencing has been put in place to safeguard members of the public while the necessary funding is secured to refurbish the bridge, and a footpath diversion remains in place.

Innovation and Technology, what discussions he has had with UKRI on their approach to supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative artificial intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, whether he has plans to commission a domestic research strategy for (a) detecting, (b) preventing and (c) removing deepfake images created by generative artificial intelligence.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, whether he has had discussions with the Science and Technology Facilities Council on their approach to supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative Artificial Intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, whether he has had discussions with Research England on their approach to supporting research into (a) detecting, (b) preventing and (c) removing child sexual abuse images created by generative Artificial Intelligence since July 2024.

The government has regular meetings with research and business organisations on matters relating to online safety and AI. The Online Safety Act places duties on platforms to proactively detect, prevent and remove child sexual abuse material (CSAM), including CSAM content created using AI technology. Ofcom has set out steps providers can take for these duties in draft codes of practice and will develop these iteratively. These steps include measures to detect, prevent and remove CSAM. The Act requires Ofcom to consult extensively when drafting its codes and Ofcom has an ongoing programme of research. The first code is due to come into force in Spring 2025.

Innovation and Technology, whether he has had discussions with the Regulatory Innovation Office on UK research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative Artificial Intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, whether he has had discussions with Innovate UK on their approach to supporting research into (a) detecting, (b) preventing and (c) removing child sexual abuse images created by generative Artificial Intelligence since July 2024.

The government has regular meetings with research and business organisations on matters relating to online safety and AI. The Online Safety Act places duties on platforms to proactively detect, prevent and remove child sexual abuse material (CSAM), including CSAM content created using AI technology. Ofcom has set out steps providers can take for these duties in draft codes of practice and will develop these iteratively. These steps include measures to detect, prevent and remove CSAM. The Act requires Ofcom to consult extensively when drafting its codes and Ofcom has an ongoing programme of research. The first code is due to come into force in Spring 2025.

Innovation and Technology, whether he has had discussions with the Economic and Social Research Council on their approach to supporting research into (a) detecting, (b) preventing and (c) removing child sexual abuse images created by generative Artificial Intelligence since July 2024.

The government has regular meetings with research and business organisations on matters relating to online safety and AI. The Online Safety Act places duties on platforms to proactively detect, prevent and remove child sexual abuse material (CSAM), including CSAM content created using AI technology. Ofcom has set out steps providers can take for these duties in draft codes of practice and will develop these iteratively. These steps include measures to detect, prevent and remove CSAM. The Act requires Ofcom to consult extensively when drafting its codes and Ofcom has an ongoing programme of research. The first code is due to come into force in Spring 2025.

Innovation and Technology, what discussions he has had with business stakeholders on the Government's approach to supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative artificial intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, what assessment he has made of potential merits of bringing forward legislative proposals to regulate the (a) detection, (b) prevention and (c) removal of deepfake images created by generative Artificial Intelligence.

Government made a clear manifesto commitment to ban the creation of sexually explicit deepfake images and we are bringing forward legislation to honour that commitment in the Crime and Policing Bill.Under the Online Safety Act, it is already a criminal offence to share or threaten to share a sexually explicit deepfake.We have designated the most harmful forms of deepfakes as priority illegal content, including child sexual exploitation and abuse and intimate image abuse. Services in scope will need to take proactive steps to prevent priority illegal content from appearing on their service and remove it quickly when it does.

Innovation and Technology, whether he has had discussions with the Government Office for Technology Transfer on their approach to supporting market innovation into (a) detecting, (b) preventing and (c) removing deepfake images created by generative Artificial Intelligence since July 2024.

The Secretary of State has not had any conversations with the Government Office for Technology Transfer (GOTT) on this topic. GOTT helps to accelerate government’s innovations towards the market to impact growth and deliver new products and services for citizens.Government, however, welcomes research on this important topic and has engaged with a wide range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of harmful and illegal generative AI content and to identify future research priorities.

Innovation and Technology, whether he has had discussions with the Science and Technology Facilities Council on their approach to supporting research into (a) detecting, (b) preventing and (c) removing child sexual abuse images created by generative Artificial Intelligence since July 2024.

The government has regular meetings with research and business organisations on matters relating to online safety and AI. The Online Safety Act places duties on platforms to proactively detect, prevent and remove child sexual abuse material (CSAM), including CSAM content created using AI technology. Ofcom has set out steps providers can take for these duties in draft codes of practice and will develop these iteratively. These steps include measures to detect, prevent and remove CSAM. The Act requires Ofcom to consult extensively when drafting its codes and Ofcom has an ongoing programme of research. The first code is due to come into force in Spring 2025.

Innovation and Technology, what discussions he has had with Research England on their approach to supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative artificial intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, what discussions he has had with the Arts and Humanities Research Council on their approach to supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative artificial intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, what discussions has he had with Horizon Europe on their approach to supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative artificial intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, what discussions he has had with the Advanced Research and Invention Agency on support for research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative artificial intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, whether he has had discussions with universities on their approach to supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative Artificial Intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, what discussions he has had with the Economic and Social Research Council on supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative artificial intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.

Innovation and Technology, whether he has had discussions with Innovate UK on their approach to supporting research into (a) detecting, (b) preventing and (c) removing deepfake images created by generative Artificial Intelligence since July 2024.

AI-generated content is captured by the Online Safety Act where it constitutes illegal content or content harmful to children on an in-scope service. We will also criminalise the creation of non-consensual sexual deepfakes through the Crime and Policing Bill. We welcome research on this important topic. DSIT co-led the Deepfake Detection Challenge with the Home Office to assess existing capabilities and identify innovative solutions to overcome the challenges of deepfakes. In addition, we have engaged with a range of stakeholders across industry, academia and civil society to understand the potential for further detection, prevention and removal of deepfake content and identify future research priorities.