In an age of information and data abundance, we not only have access to a greater number of choices, but new means of making them. Data-driven decision making through artificial intelligence (AI) is one of the newest frontiers for technological and governance innovation.
Artificial Intelligence
In 2017, the Canadian government announced $125 million in funding for a Pan-Canadian Artificial Intelligence Strategy with the goal of making Canada a global leader in its development. Further to this, Canada is also exploring applications for AI in governance systems. Notably, as part of a pilot program starting in June 2018, Citizenship and Immigration Canada is looking to use AI to perform legal research which will inform its response to immigration claims. The aim is to eventually use this technology to help determine refugee applications. While intended to improve efficiency, this technology will likely pose significant human rights implications.
AI and Transparency
AI is developed through machine learning, which involves training an AI on large data sets to perform whatever task it is intended to fulfill. Problematically, however, in forms of AI which use deep learning – where an algorithm teaches itself how to conduct a task based on training data – the decision making process is unknown to the machine’s creators. This problem is widely referred to as ‘black boxing’ – the idea that advanced AI’s underlying functionality is largely opaque and unknown.
In this sense, the issue of AI greatly underscores the necessity of open data and transparency in that it is important to know what data is being used in its development so as to better understand its function and outcomes. Since AI is increasingly being integrated into systems which determine people’s access to public and private services, it also offer individuals recourse where they may be subjected to data discrimination which impacts their digital inclusion and by extension, their social inclusion.
AI and Privacy
AI also raises privacy concerns, not only because of the massive amounts of data involved in its creation, but because its underlying algorithms can make decisions or assessments that may put people’s personal privacy and security at risk. It is exceptionally difficult to fully anonymize data and so AI may be able to not only deduce the identity of a person, but identify personal information or circumstances they may not want disclosed or may not even be aware of. In one infamous case, Target was able to determine which of their customers were pregnant and conduct targeted mail advertising through shopping data analytics – accidentally outing a pregnant teen in the process. It is easy to imagine how instances like these could threaten the security of marginalized individuals in unsupportive communities.
AI and Algorithmic Accountability
Considering these risks and challenges, much debate has arisen around the problem of algorithmic accountability – described as “the process of assigning responsibility for harm when algorithmic decision-making results in discriminatory and inequitable outcomes”. As the development and implementation of AI rapidly advances in Canada, it is increasingly pressing that we fill the current AI policy deficit and find answers to these and other integral questions.
Internet of Things
Do you use a smartwatch, a smartphone, or any other internet-enabled devices? (We are guessing yes since you are here!) Congratulations, you are a part of the Internet of Things (IoT)*. From smart cities to smart fridges, IoT is transforming how we live, work, interact, and navigate the world.
* Internet of Things (IoT): connecting physical objects such as smartwatches, vehicles, homes, appliances,
and more with access to the internet, and the ability to transfer data over a network.
Passive Data Collection
The rush to develop IoT, however, may cause more harm than good if not well managed. The reality is IoT devices are currently developing and going online at a rate that is faster than the development of security and privacy standards which govern them. From a privacy perspective, IoT can be problematic because it involves a lot of passive data collection in order to customize the service or experience for its users. The implementation of IoT in public spaces is sometimes criticized as constituting surveillance for this reason.
Security of Information
The security of IoT is also a major concern because the network is often only as strong as its weakest link. With poor security standards, it is far more likely that the data collected by IoT technologies, some of which may be personal or identifying, could be leaked and in turn compromise an individual’s privacy and security. The networked nature of IoT devices also means that an attack on one person could compromise the network as a whole. If numerous devices are hacked, they could be co-opted to launch an attack of a greater scale (e.g. a botnet attack) on the internet.
Privacy and Security by Design
For this reason, many are calling for the use of privacy and security by design* principles in the rollout of IoT. Many consultations are also ongoing to determine what the Canadian policy approach should be and how to best inform users and consumers about good practices of IoT.
* Privacy and security by design: a principle that promotes privacy and data protection compliance in the creation
of digitized devices. This principle takes a more holistic approach to the notion of privacy.
For Further Reading:
- Internet of Things (IoT)
- The Internet of Things (IoT): An Overview
- Privacy and The Internet of Things
Smart Cities
Smart cities* are one of the most popular and exciting applications for the Internet of Things (IoT) technology. The idea that an entire neighbourhood or city could be equipped with tools to improve and personally customize its visitors’ experiences in real time is nothing short of revolutionary. Many people are in support of smart cities projects because of the myriad insights they may offer to researchers, policymakers, business owners, etc. about how a city works and where it could be improved to promote sustainability, socio-economic inclusion and opportunity, as well as other indicators of wellness.
* Smart cities: a city that incorporates information and communication technologies to promote sustainability,
socio-economic inclusion and opportunity, and wellness. Smart cities collect, analyze, and use data to improve
the quality of life in an urban city.
Sidewalk Toronto
The development of smart cities is a major focus of technological innovation in Canada; however, not all feedback is positive. Sidewalk Toronto, a project which seeks to transform Toronto’s Waterfront in partnership with Alphabet’s (a.k.a. Google) Sidewalk Labs, has been the subject of much buzz – supportive and critical. Integrating smart cities technology is also a national strategic objective. In 2018, Infrastructure Canada created a Smart Cities Challenge encouraging Canadian communities of all sizes to compete for funding to implement a smart cities project.
Surveillance in smart cities
As projects like these are underway and new ones emerge, it is important to ask some serious questions about the potential implications of this technology. Many are worried about the prospect of increased surveillance enabled in smart cities and the impacts of this on privacy rights and other intersectional issues.
Marginalization in smart cities
While smart cities projects often purport to involve people-centred design, some question whether they will enrich everyone’s experience of a neighbourhood equally or lead to the elaboration of social inequalities. Digital divides (inequalities in access to technology and the internet), for instance may inhibit already marginalized groups from fully participating in their communities. Asking questions about who these projects are built for and how they may impact particular groups in our diverse society is essential if we wish to design a more inclusive future, particularly in urban settings.
Data ownership
The concept of smart cities also raises issues about data ownership. Will the data produced by the various sensors and IoT devices embedded within smart cities’ infrastructure reside in that city or country’s borders or be sent elsewhere?
If these projects are implemented in a way which undermines individuals’ rights and leads to heightened socio-economic inequalities, they are still in need of smartening up.
For Further Reading:
- Sidewalk Toronto has Yet to Give Us a Reason to Trust its Smart City Experiment
- Smart Cities Challenge
- Welcome to the Neighbourhood. Have you Read the Terms of Service?
- Sidewalk Lab’s Toronto Waterfront Tech Hub Must respect Privacy and democracy
- CIPPIC Smart Cities
- Monetizing Smart Cities