May 4, 2023

Bill C-27: Artificial Intelligence and Data Act – An Important First Step Towards a Canadian Artificial Intelligence Charter of Digital Rights and Freedoms

This month Bellissimo Law Group PC submitted its Parliamentary brief to the House of Commons Standing Committee on Industry and Technology  studying Bill C-27 the Digital Charter Implementation Act, 2022 and the Artificial Intelligence and Data Act (AIDA).[1] At the centre of it all is the explosive use of artificial intelligence (AI). Many grappling with the enormity of the implications surrounding the explosive use of artificial intelligence (AI) solutions in the world can find themselves equally in awe and in dread. Simply put, there exist incredibly positive explosive possibilities for all facets of life, let alone law.

Dread, though, derives from a fear of how much of life can be automated: what remains distinctly human, private? There are examples both domestically and internationally discussed briefly below, where AI has worsened the racial, socio-economic, and political divides and discrimination so entrenched in our society. Damaging though these outcomes may be, “they are not inevitable.”[2] Equally known, automated systems are also responsible for incredible advancements in several different sectors, from agriculture to zoology. “These tools hold the potential to redefine every part of our society and make life better for everyone.”[3]

It is important then that these developments do not negatively change democratic values and civil liberties. Thus, it is critical that any legislation related to AI encompass both the public and private sphere. One can hardly fathom what our societal and jurisprudential evolution, and its critically devalued bearing on every sector of society could have been had the Canadian Bill of Rights[4] and the Canadian Charter of Rights and Freedoms[5] applied only to private entities. It is laudable, therefore, the effort taken by the ISED to legislate the digitization of delivery, information collection, data storage, document management, and the expanding use of AI but it is nowhere near enough: we need a Canadian Charter of Digital Rights and Freedoms.

Moving forward as is, risks sending the message that AI must be legislated only in the private, not public sphere. The counter argument may be that we must begin somewhere. However, AIDA seemingly intersects between the public and private in a few of its provisions. It stands to reason that the next steps will be taken to fully explore the use of AI wherever possible, but with the benefit of safeguards implemented throughout society.

The use of AI globally explodes each year. This has not abated during the pandemic, in fact, AI has been crucial in addressing the COVID-19 pandemic[6] in many ways including the development of vaccines,[7] biomedical research and AI driven diagnostic X-Rays.[8] Montreal as one example, is now a major global hub for AI research and its AI ecosystem is thriving.[9]  Furthermore, in Estonia there are “robot judges” for small claims court for disputes valued at less than 7000 euros.[10] Digital courts with non-human judges have also been implemented in China.[11] The emergence of ChatGPT and Dall-E have dramatically raised the profile for AI uses seemingly overnight. A powerful example of the transformative properties of AI. Universities, are also beginning to revamp how they teach.[12]

The implications are boundless.  The increasing use of AI will inevitably require that interpretative, jurisdictional, and other legal issues be resolved by the Courts. Thus, no bright line, in our view, can reasonably be drawn to restrict public oversight, nor can it be left exclusively, as AIDA has been criticized for, to the government to regulate. Third parties including the judiciary must play a key role.[13]

States have begun implementing transparency and ethical guidelines of their own to end any lack of trust by citizens of the state. A prime example of this is the New Zealand Algorithm Charter that was implemented to provide guidelines on the use of AI by institutions.[14] One of its principal objectives is to provide transparency on the implementation of AI and to ensure that citizens’ rights and freedoms are upheld, despite the use of automation technologies in important decision-making processes.[15] Other governing organizations and bodies, like the European Union Commission, are similarly guided by the emergence and development of digital rights and freedoms, such as  respect for human dignity, freedom of the individual, respect for democracy, justice and the rule of law, equality, non-discrimination, and solidarity, and finally citizens’ rights (i.e., human autonomy, prevention of harm, fairness, and explicability).[16] New Zealand, Australia as another example[17], and the European Union[18], have become pioneers in the newly arising field of AI use in government institutions.  The United States has recently released the Blueprint for an AI Bill of Rights (AIBoR) which aims to serve as a response to the unrestrained potential of automated systems.[19]

However, in juxtaposition to AIDA, AIBoR is an aspirational document and, glaringly, the AIBoR opens with a disclaimer that plainly explains that it is nonbinding, especially to law enforcement. It only speaks to the Federal government and not private companies.[20] Both, AIDA and AIBoR suffer from dealing piecemeal with AI and digital regulation. A standalone strong and all-encompassing legislative enactment like we have seen outside of North America is the path forward. This is critical because AI can turbocharge discrimination.[21]

In another AI brief we filed before the Parliamentary Citizenship and Immigration Committee set out several striking examples of Canadian and international examples of bias AI driven data and its application.[22]  A few examples include the AI powered iBorderCTRL – a lie detector used by the European Union at borders – was demonstrated to discriminate against “people of colour, women, children, and people with disabilities”.[23] In New Zealand, technology used to identify potential immigrant overstayers was modelled using information such as age, country of origin, gender, usage of public health services, law enforcement encounters, and immigration status.[24] Immigration New Zealand was criticized for the use of ethnicity data in its risk modelling, as it had the potential to further marginalize racialized groups.[25] This resulted in the passing of the Algorithm Charter. Again, these issues arose when AI was utilized by government authorities.

Bill C-27/AIDA can only be seen as a first step in what must be a careful, thoughtful, and deliberate process that culminates with a Canadian Digital Charter of Rights and Freedoms that applies equally to all.  Anything else will fall short of the legal necessities of our new and still relatively unknown digital future.  AI indeed offers an enticing redefinition of large swaths of our current reality.   The law then must play a pivotal role to bridge the growing divide between the ethical and the legal and breathe new life into what our fundamental legal rights, freedoms and protections stand for in a new and seemingly daily reorientation of our existence in a digital world.

[Sources]