Law Commission Ontario: New AI Criminal Justice Framework Proposed

• The Law Commission of Ontario (LCO) has issued 29 recommendations for responsible AI use in Canada's criminal justice system.
• The LCO proposes new provincial legislation and governance frameworks for AI, including mandatory disclosure and impact assessments.
• Recommendations include amending the Enhancing Digital Security and Trust Act, 2024, or introducing new laws with a precautionary principle for AI.
• A joint Federal-Provincial committee is suggested to establish safeguards for AI evidence rules in criminal matters.
• Legal Aid Ontario is urged to create an AI Centre of Excellence and receive increased funding to support defence counsel in AI-related cases.

The report also identifies a significant regulatory void, noting that to date, no Canadian government has enacted specific laws or regulations to govern the use of AI by police or criminal courts.

The Law Commission of Ontario (LCO) has released a comprehensive final report on the integration of artificial intelligence (AI) within the criminal justice system, marking Canada's first in-depth examination of the benefits and risks associated with such technologies. This landmark document, titled "AI in Criminal Justice," puts forth 29 distinct recommendations aimed at ensuring the responsible deployment of AI by various institutions, including police forces, Crown prosecutors, and the courts. A central tenet of these recommendations is the urgent need for new provincial legislation and the establishment of robust governance and risk management frameworks specifically tailored for AI systems operating in this critical sector.

The LCO emphasizes that this proposed framework must furnish clear legal and policy directives for all stakeholders, encompassing government bodies, law enforcement agencies, judicial entities, prosecution services, defence counsel, and legal aid providers. The overarching goals are to bolster public safety, safeguard legal rights, and cultivate public trust in the justice system. Key components of this recommended framework include mandatory public disclosure of AI system usage, the requirement for dedicated impact assessments, continuous auditing and evaluation processes, and explicit guarantees to uphold judicial independence.

To solidify these governance principles, the LCO has urged the Ontario government to either amend the existing Enhancing Digital Security and Trust Act, 2024, or enact complementary legislation. Such laws would legally mandate adherence to these principles, beginning with a preamble that subjects all AI systems to a precautionary principle. Furthermore, any AI system deployed must strictly comply with the Canadian Charter of Rights and Freedoms, principles of criminal procedural fairness, privacy law, human rights law, criminal common law, and the fundamental principle of judicial independence.

The proposed legislation would also introduce mandatory risk categories for AI systems, alongside specific reporting obligations. These reports would be submitted to the province's Inspector General of Policing and Law Enforcement Complaints Agency, which would then be responsible for publishing an annual public report detailing the use of AI by law enforcement services across Ontario. This structured reporting mechanism aims to enhance transparency and accountability in the application of AI technologies.

Beyond overarching governance, the LCO's recommendations extend to the critical area of AI-generated evidence in criminal proceedings. The commission advocates for the formation of a joint Federal-Provincial Criminal AI Evidence Rules Committee. This committee, through a dedicated consultation process, would be tasked with examining and proposing fundamental safeguards for the rules governing AI evidence in criminal matters, ensuring fairness and reliability in its presentation and assessment.

Recognizing the evolving landscape for legal professionals, the LCO has also called upon Ontario's Ministry of the Attorney General to significantly increase funding and resources for Legal Aid Ontario (LAO). This enhanced support is crucial for equipping criminal defence counsel to effectively respond to and challenge cases where AI evidence is introduced. To further this objective, the commission recommends that Legal Aid Ontario establish an AI Centre of Excellence. This specialized centre would provide essential training on AI-related laws and procedures in criminal justice, set professional competence standards for AI legal research and advocacy, conduct pertinent legal research, and offer support for both individual and systemic AI test cases.

The LCO report underscores a dual reality regarding AI in the criminal justice system. While acknowledging the potential for AI technologies to enhance public safety, improve police investigations, and increase the efficiency and fairness of criminal proceedings, it simultaneously highlights significant inherent risks. Report authors Ryan Fritsch and Nye Thomas, who lead the LCO's Criminal AI Project and serve as executive director respectively, point out that systems like predictive policing, facial recognition technology, and generative AI pose proven threats to human rights, criminal procedure, privacy, and access to justice.

Crucially, these risks are not uniformly distributed, disproportionately affecting racialized and low-income communities. The report also identifies a significant regulatory void, noting that to date, no Canadian government has enacted specific laws or regulations to govern the use of AI by police or criminal courts. This absence of a legal framework, coupled with the identified benefits and acute risks, makes the LCO's recommendations a pivotal step towards establishing a responsible and equitable approach to AI integration in Canada's justice system.

Lawyers and compliance officers in Canada should monitor the proposed legislative changes and governance frameworks for AI in criminal justice, particularly regarding mandatory disclosure, impact assessments, and evidence rules. Defence counsel should anticipate new resources and training from Legal Aid Ontario to address AI evidence, while those advising law enforcement or government agencies must prepare for new compliance and reporting requirements.

Source: Original reporting via Law Commission of Ontario.