Motion #03
Science & Technology PWG: Graham Taverner*,Tim Davies, Nadine Storey, Diana Newsom, Linda Freeman, Hannah T. McLellan, George Ttoouli, Ann Mills, Diane Davies, Simon Pitts, Julie Spriddle, Paul Hoekstra, Alun Davies, Diane Davies, Jack Lenox, Celia Coram, Alan Francis, Tim Davies, Simon Pitts, Malcolm Bailey, Dagan Smith
Despite the lack of prominence given to this issue in recent election, the likely impact of AI on our lives now and in the near future demands serious and urgent action. We feel it is imperative that we have clear policy in place showing how we will regulate AI for
SOC Note: Synopsis word limit reached
The following should be added to the Science and Technology chapter of PSS:
ST 380: We recognise the potential of AI to be of great benefit, including to achieve a sustainable future for our society, but also acknowledge both immediate and evidenced harms to our freedoms, social cohesion, livelihoods and democratic processes, and concerns about existential threats. As a result, we insist that there should be comprehensive, legally-enforceable regulation of AI to address ecological, social and economic harm, and guide developments to better serve people and the planet.
ST381: We will establish a robust regulatory system to provide oversight and coordination for sectoral regulation of data and AI. A coordinating body will recruit and develop the interdisciplinary expertise required to respond promptly to new AI developments. This body’s overall aim will be to use regulation to effectively drive AI development and usage in line with the precautionary principle: ensuring systems are safe, transparent, accountable and both human and nature-centric.
We will take a risk-based approach to establish proportionate disclosure, participation, impact assessment and audit/compliance requirements for AI systems and applications. General purpose AI and generative AI pose systemic risks so will be subject to strict regulation. When risks are well managed, AI has the potential to make a positive contribution to our society, economy and environment.
ST382: We shall place deliberative and democratic processes, and the voices of those most affected by particular technology development and deployment, at the heart of data governance and AI regulation. Workers and other stakeholders must be actively involved in decisions about adoption of data and AI tools in the workplace and public life, and adoption decisions must be made at the lowest possible level in consultation with those affected.
ST383: We will introduce energy performance and water consumption labelling for AI systems to allow informed choice of their use and as a mechanism to drive improvements in their ecological impact.
ST384: Our economic and tax policies provide the basis for active intervention to address AI market concentration and monopolies, including through an internationally coordinated approach.
ST385: To ensure individuals and communities have prompt access to redress, we will create an AI Ombudsman to champion the rights of individuals and groups who submit complaints, or have complaints submitted on their behalf, relating to material or other harm from AI system output or decision-making.
ST386: Research funding, regulatory sandboxes, and public procurement processes will be used to guide AI innovation towards provision of public good, building on our strong national foundations of technical excellence, social science research and innovative business. Where AI is used to augment professional practice, we will ensure humans still make all key decisions, and where automated decision making significantly affects lives, this is done with care and humanity, and with speedy access to mechanisms for correction and redress.
ST387: We will prohibit certain uses of AI by law, including lethal autonomous weapons systems (See PSD541); AI systems that use deception (faking, impersonation etc) to manipulate behaviour or negatively impact on informed decision-making and/or democratic processes; routine biometric identification in public places; and emotion-recognition systems in workplaces or educational settings.
ST388: There can be no effective AI regulation without data governance. Therefore, we will ensure that robust data protection rules, and the inalienable moral rights of creators to have their contribution to creative output recognised, are respected. Developers of AI systems must respect all applicable copyright laws. In line with policy
ST223 we recognise the value of making publicly-funded digital content freely accessible, including for training machine-learning models, alongside developing transparency standards so that sources of data may be verified and accurately referenced.
ST89: We will implement support for those whose livelihoods are disrupted by AI deployment, and will ensure workers’ rights and interests are respected when AI leads to significant changes in working conditions (See WR104 and WR206). We see Universal Basic Income (EC402) as one important tool to proactively and positively respond to potential AI impacts on working lives; we see public education in good use of AI as another tool to enable individuals the freedom to live a worthwhile life, devoting time in whatever proportions they choose to the following “seven C’s”: - Curiosity, Conservation, Challenge, Creativity, Community, Charity, Care.
ST390: Our AI regulatory system will foster international links and encourage global collaboration towards establishing a world-wide body to oversee and police AI development, including establishing norms for AI to produce output that is in harmony with established human rights principles and sustainable development goals.
Last updated on 2024-11-02 at 12:44