AI Ethics in Law: ABA Opinion 512, the EU AI Act and What Lawyers Must Do
A practical guide to AI ethics in law: ABA Opinion 512, the EU AI Act, UNESCO principles, and the duties of competence, confidentiality and supervision lawyers must uphold in 2026.
Artificial intelligence is rapidly transforming how legal services are delivered. But with that transformation comes a responsibility that every practitioner, firm, and institution must take seriously: ensuring that AI is used ethically, accountably, and in a manner consistent with the duties lawyers owe to their clients and to the courts.
International Guidelines Setting the Standard
Several leading international bodies have issued frameworks that directly shape how AI should be used in professional legal contexts.
The United Nations has adopted ten core principles for the ethical use of AI across all UN system entities, grounded in human rights and ethics. These include: do no harm; defined purpose, necessity and proportionality; safety and security; fairness and non-discrimination; sustainability; right to privacy and data governance; human autonomy and oversight; transparency and explainability; responsibility and accountability; and inclusion and participation.
The European Union's High-Level Expert Group on AI published its Ethics Guidelines for Trustworthy AI, setting out that trustworthy AI must be lawful, ethical, and robust. The EU framework identifies four core ethical principles: respect for human autonomy; prevention of harm; fairness; and explicability.
UNESCO operates its Judges Initiative in over 160 countries, training judicial operators to apply international human rights standards to AI-related challenges including bias, discrimination, privacy, and transparency.
The American Bar Association (ABA) issued Formal Opinion 512 in July 2024 — its first formal ethics opinion on generative AI — confirming that lawyers using AI must fully consider their ethical obligations under the Model Rules of Professional Conduct. The Opinion covers six key ethical dimensions: competence, confidentiality, communication, candour toward the tribunal, supervisory responsibilities, and fees.
The Core Ethical Challenges
Competence and Verification
AI outputs — particularly from generative tools — can be confidently wrong. The ABA has made clear that a lawyer's uncritical reliance on AI output without appropriate independent verification may constitute a breach of the duty of competence.
Confidentiality and Data Privacy
Lawyers handle privileged, sensitive, and client-confidential information as a matter of course. Uploading such materials to a general-purpose AI tool — where data may be used for model training — creates a direct conflict with the duty of confidentiality. The ABA's Opinion 512 explicitly addresses this risk, requiring lawyers to evaluate the data-handling practices of any AI tool before using it with client information.
Candour Toward the Tribunal
The hallucination problem in AI is not merely an inconvenience — in a legal context, it is an ethical crisis. Lawyers have a duty of candour to courts under rules such as ABA Model Rules 3.1, 3.3, and 8.4(c). Submitting AI-generated content containing fabricated citations or misrepresented law to a court is not just embarrassing; it may be a sanctionable ethical violation.
Supervisory Responsibility
The ABA Opinion requires managerial lawyers to establish clear policies on permissible AI use, and supervisory lawyers to ensure that all staff — including non-lawyers — are trained in the ethical and practical use of AI tools. This obligation extends to work outsourced to third parties who use AI in their processes.
A Path Forward
Ethical AI adoption in legal practice is not about avoiding AI — it is about deploying it responsibly.
Firms that invest in purpose-built legal AI tools, establish clear usage policies, and train their people to verify and supervise AI outputs will be best positioned to harness AI's genuine benefits while honouring the duties that define the profession.