AI 및 기술법
인공지능 규제, 알고리즘 책임 및 자동화된 의사결정.
소개
AI 법률은 인공지능과 신기술이 제기하는 법적 도전을 다룹니다.
In Brief
There is no single UK AI Act; AI is regulated through existing frameworks. The UK GDPR restricts solely automated decision-making with legal or significant effects (Article 22). AI systems producing discriminatory outcomes may breach the Equality Act 2010. Product liability for AI-caused harm may arise under the Consumer Protection Act 1987. Copyright in AI-generated works is uncertain (CDPA 1988, s.9(3)). The AI Safety Institute evaluates frontier AI models. Ofcom regulates the Online Safety Act 2023 duties for online platforms.
핵심 원칙
UK GDPR & Automated Decision-Making — Article 22 restricts solely automated decision-making that produces legal or significant effects. Data subjects have the right not to be subject to such decisions and to obtain meaningful information about the logic involved.
Algorithmic Transparency — The UK Government's Algorithmic Transparency Standard requires public sector organisations to publish information about algorithmic tools used in decision-making.
AI Safety — The AI Safety Institute (established 2023) conducts evaluations of frontier AI models. The UK approach is 'pro-innovation' relying on existing regulators rather than new AI-specific legislation.
Intellectual Property & AI — Copyright in AI-generated works is uncertain. Under s.9(3) CDPA 1988, copyright in computer-generated works belongs to the person who made the arrangements for creation. The scope of this provision for modern generative AI is unclear.
Product Liability — AI systems causing harm may engage product liability under the Consumer Protection Act 1987, negligence under Donoghue v Stevenson principles, or sector-specific regulation.
Bias & Discrimination — AI systems that produce discriminatory outcomes may violate the Equality Act 2010. The ICO and EHRC have issued guidance on algorithmic fairness.
Online Safety Act 2023 — Imposes duties on platforms regarding illegal content, content harmful to children, and user empowerment, with Ofcom as the regulator.
핵심 법령
주요 판례
Thaler v Comptroller-General
[2023] UKSC 49
Lloyd v Google
[2021] UKSC 50
일반적인 시나리오
AI hiring tool discriminates against women
An AI recruitment tool that systematically disadvantages candidates sharing a protected characteristic may constitute indirect discrimination under the Equality Act 2010, s.19. The employer remains liable even if the bias is unintentional.
Autonomous vehicle causes injury
The Automated Vehicles Act 2024 establishes liability framework. The insurer of an autonomous vehicle is primarily liable for accidents where the vehicle is driving itself, with recovery rights against the manufacturer.
AI generates defamatory content
Liability for AI-generated defamatory statements may fall on the publisher (platform), the developer, or the person who directed the output. The Defamation Act 2013 and intermediary liability rules under the Online Safety Act 2023 are relevant.
Related Careers
Frequently Asked Questions
Can an employer use AI to make decisions about hiring or firing in the UK?
Employers can use AI tools in recruitment and HR processes, but solely automated decision-making that produces legal or significant effects on individuals is restricted under Article 22 UK GDPR. Individuals have the right to request human review of such decisions, to express their point of view, and to contest the outcome. Where an AI system produces discriminatory outcomes (for example, disadvantaging candidates by gender or race), it may also engage the Equality Act 2010.
Who owns the copyright in content created by an AI in the UK?
Under s.9(3) of the Copyright, Designs and Patents Act 1988, copyright in a computer-generated work belongs to the person who made the necessary arrangements for the creation of the work. This provision predates modern generative AI and its application is legally uncertain. The Supreme Court in Thaler v Comptroller-General [2023] UKSC 49 refused patent rights to an AI inventor, and the UK Intellectual Property Office is consulting on reform. Practically, copyright in AI-assisted works may rest with the prompter or developer depending on the level of human creative input.
Is the UK subject to the EU AI Act?
No. The UK left the EU before the EU AI Act came into force and is not bound by it. The UK Government has taken a 'pro-innovation' approach to AI regulation, relying on existing sector regulators (such as the FCA, ICO, Ofcom, and CQC) to apply AI-specific guidance within their domains rather than creating a single overarching AI statute. The AI Safety Institute (established 2023) focuses on frontier AI model evaluation.
What legal protection is there against deepfake images in the UK?
The Online Safety Act 2023 created a new criminal offence of sharing intimate deepfake images without consent. The Criminal Justice Act 2003 and existing harassment legislation may also apply. Additionally, creating or sharing a deepfake for the purpose of fraud or blackmail may engage existing criminal offences. The government has indicated further legislation on deepfakes more broadly is under consideration.
Important Deadlines
Typical Costs
Official Resources
What To Do Next
Step-by-Step Guides
Know Your Rights
Common Scenarios
Get Professional Help