When artificial intelligence makes a mistake, the responsibility doesn't disappear into the algorithm,it lands squarely on the humans and organizations using the tool. A growing body of legal cases is making this principle crystal clear, with courts rejecting the notion that AI can serve as a scapegoat for inaccurate information, biased decisions, or copyright violations.

What Happens When AI Gives You Bad Advice?

One of the most widely reported examples involved Air Canada's customer service chatbot. A passenger relied on information the chatbot provided about bereavement fares and purchased flights accordingly. When the airline later refused to honor the advice, Air Canada argued that the chatbot was essentially responsible for its own statements. A British Columbia tribunal rejected that argument entirely and held Air Canada responsible for the inaccurate information published through its own system.

The lesson extends far beyond airlines. In a recent case involving the law firm Pinsent Masons, a junior solicitor used AI to draft letters containing an incorrect legal proposition. The AI had actually warned the user to verify the authorities, but this verification step was skipped. Supervisors also failed to check the output before it was sent to court. The judge described the failures as serious and emphasized that lawyers remain responsible for checking the accuracy of material generated using AI tools.

How Can Organizations Protect Themselves From AI Liability?

• Verify AI Output: Never assume AI-generated content is accurate. Whether the tool produces reports, customer communications, or legal documents, human review must happen before publication or submission.
• Maintain Professional Judgment: AI can be a powerful assistant, but professional judgment cannot be delegated to software. Supervisors and decision-makers remain accountable for what goes out under their organization's name.
• Audit Training Data: Understand what data your AI system learned from and whether that data contains bias. If historical data reflects discrimination or skewed patterns, the AI may replicate and amplify those problems.
• Document Oversight Processes: Establish clear procedures for reviewing AI outputs before they're used in customer-facing communications, legal filings, hiring decisions, or other high-stakes applications.

The Pinsent Masons case illustrates why oversight matters. The firm referred itself to the SRA (Solicitors Regulation Authority) after the incident, acknowledging that responsibility for accuracy cannot be outsourced to technology.

Can AI Bias Lead to Legal Liability?

Amazon's recruitment tool offers a cautionary tale about what happens when AI learns from biased data. The company developed an AI-powered recruitment system intended to identify the best candidates for technical roles. However, because the system was trained using historical recruitment data that predominantly reflected successful male applicants, it learned to favor male candidates. Reports indicated that the software downgraded CVs containing references to women's organizations and all-women colleges. Amazon ultimately abandoned the project, highlighting a central challenge of AI: if the data contains bias, the technology may replicate and amplify it.

This pattern raises serious legal exposure. Discrimination in hiring violates employment law, and organizations cannot escape liability by claiming the algorithm made the decision. The human decision to deploy an untested system without auditing its outputs for bias remains the organization's responsibility.

What About Copyright and AI Training Data?

One of the most significant legal battles involves Getty Images and Stability AI. Getty Images alleges that Stability AI has used millions of Getty's copyrighted images, without permission, to train its AI image generator, Stable Diffusion. The decision is currently under appeal and remains an important case in the development of AI and intellectual property law.

This case demonstrates that the biggest legal risk for many businesses may not be what AI creates, but what it learned from and whether it had the right to learn from it. As courts worldwide grapple with these questions, the principle is becoming consistent: organizations cannot simply claim ignorance about their AI system's training data or outputs.

While many of the headline cases have emerged overseas, UK courts, regulators, and professional bodies are increasingly grappling with the same issues. Whether the issue is discrimination, inaccurate advice, misuse of personal data, or copyright infringement, the law is unlikely to accept "the AI did it" as a satisfactory defense.

The future of work may involve AI, but the future of accountability remains human. Organizations embracing AI tools need to understand that deployment without proper oversight, verification, and bias auditing creates legal exposure. The courts have spoken, and the message is clear: responsibility cannot be delegated to software.