What is an AI Hype Cycle and How Is it Relevant to Canadian Immigration Law?

Recently I have been reading and learning more about AI Hype Cycles.

I first learned this term from Professor Kristen Thomasen when she did a guest lecture for our Legal Methodologies graduate class and discussed it with respect to her own research on drone technology and writing/researching during hype cycles. Since then, in almost AI-related seminar I have attended the term has come up with respect to the current buzz and attention being paid to AI. For example, Timnit Gebru in her talk for the GC Data Conference which I recently attended noted that a lot of what is being repackaged as new AI today was the same work in ‘big data’ that she studied many years back. For my own research, it is important to understand hype cycles to ground my research into more principled and foundational approaches so that I can write and explore the changes in technology while doing slow scholarship notwithstanding changing public discourse and the respective legislative/regulatory changes that might follow.

A good starting point for understanding hype cycles, especially in the AI market, is the Gartner Hype Cycle. Who those who have not heard the term yet, I would recommend checking out the following video:

Gartner reviews technological hype cycles through five phases: (1) innovation trigger; (2) peak of inflated expectations; (3) trough of disillusionment; (4) slope of enlightenment, and plateau of productivity.

It is interesting to see how Gartner has labelled the current cycles:

One of the most surprising things to me on first view is how automatic systems and deicsion intelligence is still on the innovation trigger – early phase on the hype cycle. The other is how many different types of AI technology are on the hype cycle and how many the general public actually know/engage with. I would suggest at most 50% of this list is in the vocabulary and use of even the most educated folks. I also find that from a laypersons perspective (which I consider myself on AI), challenges in classifying whether certain AI concepts fit one category or another or are a hybrid. This means AI societal knowledge is low and even for some of the items that are purportedly on the Slope of Enlightment or Plateau of Productivity.

It is important to note before I move on that that the AI Hype Cycle also has been used in terms outside of the Gartner definition, more in a more criticial sense of technologies that are in a ‘hype’ phase that will eventually ebb and flow. A great article on this and how it affects AI definitions is the piece by Eric Siegel in the Harvard Business Review how the hype around Supervised Machine Learning has been rebranded into a hype around AI and has been spun into this push for Artificial General Intelligence that may or may not be achievable.

 

Relevance to the Immigration Law Space

The hype cycle is relevant to Canadian immigration law in a variety of ways.

First, on the face, Gartner is a contracting partner of IRCC which means they are probably bringing in the hype cycle into their work and their advice to them.

Second, it brings into question again how much AI-based automated decision-making systems (ADM) is still in the beginning of the hype cycle. It make sense utilizing this framework to understand why these systems are being so heralded by Government in their policy guides and presentation, but also that there could be a peak of inflated expectations on the horizon that may lead to more hybrid decision-making or perhaps a step back from use.

The other question is about whether we are (and I am a primary perpetrator of this) overly-focused on automated-decision making systems without considering the larger AI supply chain that will likely interact. Jennifer Cobbe et al talk about this in their paper “Understanding accountability in algorithmic supply chains” which was assigned for reading in my Accountable Computer Systems course. Not only are there different AI components, providers, downstream/upstream uses, and actors that may be involved in the AI development and application process.

Using immigration as an example, there may be one third-party SAAS that checks photos, another software using black-box AI may engage in facial recognition, and ultimately, internal software that does machine-learning triaging or automation of refusal notes generation. The question of how we hold these systems and their outputs accountable will be important, especially if various components of the system are on different stages of the hype cycle or not disclosed in the final decision to the end user (or immigration applicant).

Third, I think that the idea of hype cycles is very relevant to my many brave colleagues who are investing their time and energy into building their own AI tools or implementing sofware solutions for private sector applicants. The hype cycle may give some guidance as to the innovation they are trying to bring and the timeframe they have to make a splash into the market. Furthermore, immigration (as a dynamic and rapidly changing area of law) and immigrants (as perhaps needing different considerations with respect to technological use, access, or norms) may have their own considerations that may alter Gartner’s timelines.

It will be very interesting to continue to monitor how AI hype cycles drive both private and public innovation in this emerging space of technologies that will significantly impact migrant lives.

Part 2B – An Annotated Review of Li and the Unforeseen and Unsettled Legal Consequences of Expanding the Definition of Espionage

 

 

Welcome back folks!

I had a bit of a busy several weeks since my last post – as I am taking an accountable computer systems course, learning about encryption, block chain, TOR and all the cool things I wish I knew earlier!

I have not forgotten about the Li decision. I will admit I have lost sleep over it, been confused over it, and had numerous client consultations over it. The recent development of the Named Research Organizations list, although in a much different context, have started to shed light on what institutions may be targetted and flagged. I presume many of these institutions (if not all) are risk indicators in the Integrity Trends Analysis Tool and may trigger the automation of the Security Screening Automation process.

Also, I should be on a podcast with Steven Meurrens and Deanna Okun-Nachoff talking about this decision shortly. I think it will be a fascinating conversation. Will share link!

For the purposes of this blog, however, let us jump back into the Federal Court’s decision in Li v. Canada (MCI) 2023 FC 1753 to get to the heart of the Chief Justice’s analysis.

Let’s start now with VI. Issues at paragraph 24. I do not yet have the benefit of the parties submissions to determination how the issues were framed in factums. Based on what the Chief Justice writes later in his VIII. Analysis at para 29, it seems like this question was framed by the Applicant.

Nevertheless, I think there might be some tension in the framing of the issues and then the setting out of the standard of review in VII. Standard of Review where the Chief Justice re-iterates that the Court’s limited role within the judicial context, the introduction the case itself, and the eventual function of attempting to carve out a definition for espionage.

I will note that this is not rare, however. We have seen it in many contexts, and indeed the Chief Justice has also engaged in a similar discussion of the role of a comparative approach in the s.25 H&C test in Huang v. Canada (MCI), 2019 FC 265.

Finally, for the purposes of this blog to keep it shorter I will focus only on paragraphs 29 to 50 and leave for the next blog the “Application to the Decision” section.

Moving to the Analysis in Section VIII.

The first issue is whether the Officer erred in applying an overly broad term ‘espionage’ under s.34(1)(a) IRPA¬†(see para 29). The Chief Justice notes that there is no definition of the term “espionage” in¬†IRPA,¬†or it appears, in any Act of Parliament. This is crucial because I think it highlights a clear legislative/policy gap that IRCC will need to look to fill.

There are some legislation that engage in for example a definition of economic espionage in the Security of Information Act, but the context of the act and who it has been used to prosecute does make it very different and difficult to translate to the immigration setting. I see this omission as an emerging gap for legislators to step in.

There are two key paragraphs in Li involving the definition of espionage, that frame the decision. The Chief Justice writes at paragraph 31 and 32:

[31]¬†However, Mr. Li submits that the term¬†‚Äúespionage‚Ä̬†has the following five characteristics:

(1) There is an aspect of secrecy, clandestineness, surreptitiousness, or covertness in the way the information in question is gathered.

(2) The information is collected without the other parties’ knowledge and consent.

(3) The collector, by the time they are actively engaging in information gathering, does so under the control and direction of a foreign entity.

(4) The information is regarded as secretive, as opposed to simply private.

(5) The act is against Canada or contrary to Canada’s interests.

[32]¬†I disagree. In my view, the jurisprudence supports a broader definition of¬†‚Äúespionage.‚Ä̬†At its most basic level, the concept of¬†‚Äúespionage‚Ä̬†contemplates the secret, clandestine, surreptitious or covert gathering¬†or¬†reporting of information to a foreign state or other foreign entity or person. When such activity is against Canada or is contrary to Canada‚Äôs interests, it falls within the purview of paragraph 34(1)(a).

There are several complications created by the definition generating process: (1) what constitues reporting? (2) what constitutes information? (3) what constitutes a foreign entity? (think of foreign-controlled companies operating in Canada for example) (4) who is a foreign person? (is it entirely immigration-status related or more than that?) (5) We also return back again to what are Canadian interests and are the relevant times of when actions occured and interests considered material?

Also, by way of the way it is gramatically structured does the reporting of information to a foreign entity/person have any modifier. It appears in the Chief Justice’s decision it can be public information, but surely the gathering or reporting of any public information to a foreign person would be an overbroad definition.

The Chief Justice summarizes at paragraph 47:

[47]¬†In summary, and having regard to the foregoing, I consider that the term¬†‚Äúespionage‚Ä̬†contemplates (i) the secret, clandestine, surreptitious or covert gathering of information on behalf of a foreign government or other foreign entity or person,¬†or¬†(ii) the reporting or communication of information, whether surreptitiously or publicly gathered, to such a recipient. I further consider it reasonable to include within the definition of¬†‚Äúespionage‚Ä̬†the unauthorized reporting or communication of such information to a third party acting as an intermediary for the transmission of the information to such a recipient. When such activity is against Canada or is contrary to Canada‚Äôs interests, it falls within the purview of paragraph 34(1)(a). This is so even if the information in question was gathered in public.

This is interesting as it then adds a modification of unauthorized, but is it a necessary condition. How does one seek authorization? Does it have to be in writing or could it be oral? If knowledge and consent is provided, is this information not authorized for disclosure?

It seems like the words on behalf of have significant play but does it apply to only foreign governments or as well entities or persons.

I think we will also eventually need to get some clarity as to what ‘such a recipient‘ means.

For example, if a permanent resident or international student goes home from a day of work to discuss a public university research project/grant with their foreign national spouse they are working on would that constitute espionage if the information transmitted potentially contrary to Canada’s interests? For example, if the spouse asks how much money the project is worth financially and how much they will get paid, could that constitute espionage under a specific fact patter?

What if the information being collected or gathered is on behalf of themselves but at risk of disclosure in the future (either intentionally or not) to a foreign entity or person that may benefit that entity or indiviudal. What if it is written in a resume or spoken of in a job interview with a potential foreign employer?

If a journalist is a foreign investigative correspondent paid by a foreign entity is looking into the Canadian Government’s international policy through publicly accessible ATIP information would that constitute espionage?

The only thing linking it all would be the act being Contrary to Canada’s interest and requiring some sort of intent to actually gather the information.

As the Chief Justice writes at paragraph 48.

It will suffice if that information, even if publicly available, was communicated or reported upon to a foreign state or other foreign entity or person, without any authorization. 

This suggests that a lack of authorization is a key part of an espionage test, and that the other modifiers of secrecy, clandestineness, surreptitiousness, or covertness are not needed, that neither is control or direction of a foreign entity or person, nor the lack of knowledge or consent (para 48).

I will summize that I am not certain what constitutes espionage after reading this section of the case. For one, I think commas, subsections, and a list are needed for want of misinterpretation or incorrect reading.

I also wonder – had I been a permanent resident or foreign national (and not a Canadian citizen), whether my own advice to my foreign national clients, based on the information I have gathered from my investigative research of Canadian immigration practice and policy (publicly available information), might constitute espionage. I received releases from ATIP which were releasable to me, but certainly not giving me a broad authorization to share online and have it read potentially by foreign entities or persons. I use this information in the interests of access to justice and to critique the system I work in and hopefully transform it for public good. It is, however, arguably contrary to Canada’s interests to have knowledge of things such as Chinook and triage made public?

What about my colleagues who practice in immigration who are permanent residents? Are they committing espionage by advising their clients utilizing information they have gathered through ATIP and information requests?

By writing blogs and sharing them on online platforms am I communicating and reporting? What about posting a video on TikTok or WeChat or another foreign state-owned entity – is this considered communicating and reporting? What if one of my clients were a foreign entity or individual in a country deemed hostile by Canada?

Perhaps I am confused and missing the boat but I feel like I cannot competently advise a client right now Рfor example Рon whether or not to try and seek entry into Canada at this time in the event any past, current, or future action they may take could be deemed espionage.

Did it matter that Li came to study? What if he came to work at McDonalds? What if he came to see a loved one in Canada for two weeks? Is he still at risk to gather information and pass it on to a foreign entity just by his very presence in Canada?

While I will save the heart of my analysis on the reasonable grounds to believe standard to the next blog part where the Chief Justice looks at the application of the law to the facts […]

Part 2A – An Annotated Review of Li and the Unforeseen and Unsettled Legal Consequences of Expanding the Definition of Espionage

 

Introduction

As promised, it is time for part II of my blog part series on the Federal Court decision of Li v. Canada (Citizenship and Immigration) 2023 FC 1753.

I will write this blog over several days. Today represents Part 2A which covers Sections I-V of Chief Justice Crampton’s decision.¬† Sections VI to VIII, which includes the issues and analysis, will form Part 2B. To keep this more accessible to a more general audience and given the broad implications of this decision, I will try my best to keep this as plain language as possible.

Today, I set the scene a bit with a lot of interesting preliminary discussions and factual/legal framing, in the next I engage the core of the legal analysis with a review of the issues, standard of review, and analysis sections of the decision (aforementioned Part 2B) and if you stay until third part (Part 2C) I will highlight some of the unforeseen and unsettled legal consequences created by the decision. This decision is simply a gamechanger for Canadian immigration law as we head to uncertain times.

I want to be clear at the outset that what I will also focus more on the substantive nature of the security regime and inadmissibility, rather than to try and analyze the judgment from a purely administrative law lens of fairness and reasonableness. As such, my concerns too are going to be centered and focused on the uncertainties created by an inadmissibility regime that punishes individuals not necessarily for what they have done, but for what they may do – and my call for a greater personalized and individualized assessment needed for such a finding to be made, given the severe consequences of being labelled as an individual inadmissible for espionage.

As a further prelude, I will say that what I have noticed from the Chief Justice’s last two major decision, Li and his decision in Sidhu,¬†involving the horrible Humboldt Broncos tragedy, shows a willingness to engage in the broader societal impacts of immigration’s administrative law consequences. I am aware that the triage system for selecting cases involves the Chief Justice choosing the assignment of certain cases among the judges. It is not surprising in my mind that he chose these two cases to take on, rendered in close succession, that have generated significant outside attention.

I suspect administrative law will receive more of this “public” attention moving forward, and will be asked to interrogate larger societal questions – involving issues such as racism, bias, technological developments, inadmissibility, Indigenous sovereignty, among other hot button issues. As the Federal Court becomes more accessible and even more relied upon, folks will pay more attention. Decisions that are more responsive, written for losing parties, and aware of the potential consequences of either trying to establish or avoid establishing precedent/precedential value will be very crucial.

Now without further ado, let’s get into Li.

 

The Li Decision

I. Overview

Similarly to Sidhu, the Chief Justice starts the decision off in paragraph 1 with quite a bold statement. After I read it the first time, I both knew this decision would be impactful but also had a gut sense before even reading the facts that this probably was a decision favouring the Government.

[1] As hostile state actors increasingly make use of non-traditional methods to obtain sensitive information in Canada or abroad, contrary to Canada’s interests, the Court’s appreciation of what constitutes espionage must evolve.

A couple things to note in this first paragraph.

First, the word¬†hostile¬†definitely raises flags. One asks what countries are currently hostile? What is the timeframe considered for the hostility. One also thinks of about Canada’s interests. A few years ago Canada’s interests were apparently economic trade-driven with certain countries. Those interests could change depending on window. In the criminality and equivalency context, Tran advised us about retroactivity/retrospectivity and ensuring alignment of individuals knowing the consequences at the time they commit an action. I think the national security context arguably skews this context signficantly, but here we have now seen individuals who are punished (in a non-criminal sense) for associations they may have had in the past and tying these to possibly forseeable future events that may occur in the future, without¬†even having committed any action per se.¬†We have seen cases like¬†Geng, from last year, where individuals who were once permanent residents of Canada having cleared security checks from years prior are being re-engaged by the systems as the investigative goalposts and geo-politics have shifted.

A reminder and as a framing point, this idea of “Canada’s interests” is from the legislation itself in s.34(1)(a) of IRPA.

Security

  • ¬†(1)¬†A permanent resident or a foreign national is inadmissible on security grounds for

    • (a)¬†engaging in an act of espionage that is against Canada or that is contrary to Canada‚Äôs interests;

Second, it is quite telling that the Chief Justice utilized the wording “Court’s appreciation.” To me it represents, and quite correctly so, at least a stated intention to not cross over to stepping into the role of the legislature or to re-litigate the case. He wants to portray this is a case about judicial intepretation. We can assess later how well the decision reflects this, in situ.

The following three paragraphs complete the overview, including summarizing the Applicant’s arguments – namely that the Officer adopted an overly broad definition of the word “espionage” and that evidence was misapprehended and ignored, and stating his decision to reject the application (paras 4-5).

II. Background

Starting at paragraph 6 we start to learn more aobut Mr. Li. He is a PRC citizen. He attempted to apply to the University of Waterloo for a PhD Program in Mechanical and Mechatronics Engineering. There were delays in background checks (a common issue I have recently commented about). The Applicant was given a last extension to obtain a study permit for a PhD project (para 7). This last extension nature may have eventually become a double-edged sword when it came to the matter covering into a JR of the final decision, but also in the parsing of a non-need to certify a question (as we will discuss in the next blog). We learn through the judgment as well that the decision was started by way of what was likely mandamus (para 8).

We learn also that the Minister applied for non-disclosure of certain information in the Certified Tribunal Record (“CTR”) under s.87 of the IRPA. Let us pause here to take a look at that provision.

Application for non-disclosure ‚ÄĒ judicial review and appeal

¬†The Minister may, during a judicial review, apply for the non-disclosure of information or other evidence. Section 83 ‚ÄĒ other than the obligations to appoint a special advocate and to provide a summary ‚ÄĒ applies in respect of the proceeding and in respect of any appeal of a decision made in the proceeding, with any necessary modifications.

2001, c. 27, s. 87

2008, c. 3, s. 4

2015, c. 20, s. 60

For those interested in another s.87 redaction case where the matter got a bit more complex with the applicant having actually succeeded in removing the redactions, check out Kiss v. Canada (Citizenship and Immigration), 2023 FC 1147 (CanLII), <https://canlii.ca/t/jzwtx> at paras 21-34.

We learn that Mr. Li learns from this redacted certified tribunal record (“CTR”) that the Center for Immigration National Security Screening recommended that there are reasonable grounds to believe he is inadmissible under s.34(1)(a) of IRPA.

Another pause. What is the Center for Immigration National Security Screening? I will not go into too much detail here (a whole other blog topic) but for some light background reading I would suggest looking at the “Evaluation of the Immigration National Security Screening Program” posted by the CBSA.

I also have knowledge that they are utilizing technological automation in these cases through the Security Screening Automation (“SSA”) project, per the unreleased draft Algorithmic Impact Assessment (“AIA”).

 

What we likely think happened, and we know of several other institutions that have been tagged with risk indicators (using tools such as the Integrity Trends Analysis Tool).

Paragraphs 11 and 13 of the decision then provide some interesting context. We learn that the Respondent represented that it would not rely on redacted information for the purposes for the purpose of responding to the application for judicial review, but also that the Officer did not rely on any redacted information in making the Decision.

I am still awaiting a copy of the file record from the Federal Court, but I do question, especially getting the information we did about the risk indicators in Kiss through this preliminary decision on the Minister’s s.87 motion how it could not have been relied on in some way. Presumably, the redacted information was indicator information, showing how the particular institution was flagged that led to the investigation. How the school (Beihang, we learn in paragraph 15) was flagged, what information was provided to the flaggers, and the technology utilized is something I forsee will be a point of legal conflict moving forward.

The other point to take from this section is the mandamus application, it appears triggered the steps taken and in this case the Chief Justice actually ordered a decision to be made within three weeks (see para 12).

While I have had mandamus claims trigger negative action (a concern that is often, in my opinion, under-discussed as a possibility), I have not yet to date had the Federal Court judge direct a decision to judgement within one of my proceedings. What we learn is that this led to a procedural fairness interview four days before the deadline, and it appears a refusal shortly thereafter.

I question whether there was an opportunity (beyond an interview) for the Applicant to put in evidence, such as expert affidavits, to counter the Government’s position. However, I will note – based on my own experience as counsel – that attempting to gather evidence […]

International Students Making Refugee Claims: How Data + Chinook Might Meet

As a recent tweet from Steven Meurrens shows, the relationship between refugee claimants and international students is one that IRCC actively tracks.

Thanks to an Access to Information and Privacy (“ATIP”) request received earlier this year, we finally have our first look into the question of how this data might be operationalized and related to the use of the Chinook’s application processing system.

The example, we have is from January 2021 in a document titled “By the Books: Analysis of Sri Lankan Student Claimants” from IRCC’s Migration Integrity, Integrity Risk Management Branch. The document is highly redacted so we can only share what we know and speculate about what the document might otherwise also say. I have provide the document for review below

Pages from A-2021-18633AdditionalRelease (003) ‚Äď Sri Lankan Students

What the Document Says

The first part of the document summarizes four key takeaways, with part of the first and last two being redacted. Based on what we see, the summary captures the change in asylum claimants alongside changes in temporary residence and temporary resident visa issuance. This document coinciding with the pandemic, this also factors in. It is likely that the missing points pertain specifically to study permit holders, given the nature of the document and the redactions.

The Background section then delves into a bit of a country-conditions summary of Sri Lanka coinciding with entry as a top 20 source country for asylum claims in Canada. Those who practice in refugee law, will draw similarities to some of the reports found in National Documeptatin packages here. One of the pinpoints of this particular document is of the Tamil diaspora, tying into the entry of Sri Lankan nationals in 2009-2010 from the much documented Ocean Lady and Sun Sea marine vessels.

Keeping in mind this was nearly three years ago, the amount of data in this document is staggering Рfrom tracking the mode of arrival (air versus other methods), versus the documentation of those with status in Canada. The document provides a graph comparing the number of TR approvals and asylum claims before pinpointing students as having the highest claim rate among TR business lines from Sri Lanka.

The data even tracks the time frame from TR (temporary resident) issuance to Asylum Claim date, showing the Government tracking that a majoirty of students intendes to claim when they acquired their study permit. The data further delves into what level of study the students are in. In a redacted section, this data also goes into what select educational institutions they are coming from, noting over half are at select universities.

Based on an unredacted foot note, there is a reference to “Fraudulently obtained S-1 TRVs for Cape Breton University by CBSA, November 2019.”

This data is then combined with nom-compliance and IRCC’s compliance reporting history, aligning with key indicators of potential non-compliance. While it is redacted, it would be interested to see how they tracked compliance alongside actual claims made, given once an individual makes a claim they acknowledge their inadmissibility and be motivated to discontinue studies. The redactions in this section make it difficult to parse this question of whether non-compliant studies led to claims, or claims led to non-compliant studies.

A fully redacted section called “Additional Observation: Address Clustering” presumably talks about the cities in which these claimants are living. This is another further factor that is likely built into the recommendation below.

The Recommendation of Chinook Module 5 Indicators

The Next Steps section is nearly fully redacted, but the first discussed step is not and is very telling in what it states:

Immediate Actions

  1. Creation and Distribution of Sri Lankan Student Indicators

Action: The MIT recommends the construction of indicators for use by visa processing officers in Chinook Module 5. Similar to indicators used in other lines of buisness in other source countries for claims, these indicators will assist officers in identifying potential high risk cases in the Sri Lankan cohort.

Based on the above document, one would suggest that the possible redactions might even pinpoint what those risk indicators look like. I am wondering also how the risk indicators would pick up an individual as Tamil (likely through language), but then also layer on the educational institutional they are attending, the city they are living in (the pockets) and possibly other factors that are discussed in the redacted sections.

 

Implications: Understanding How Risk Indicators are Created

While we have known for awhile that refugee claims are consideredadverse outcomes for international students, this document truly challenges the breadth and scope of the type of data being used to back the data-based systems within Chinook’s Module 5 Risk Indicators. It suggests that there are data-based calculations and equations being drawn and made of all applicants and form factors outside of an Applicant’s, and currently a reviewing court’s, control.

We have written about risk indicators in the past:

Why If There’s No “N/A” Risk Flag on Your GCMS Notes, You May Have Been Risk Flagged

In this blog, we talk about how these indicators feature prominently in the spreadsheet presentation used by Officers to determine cases.

We’ve also shared that these risk indicators are being flagged through automation and AI on files (see: Integrity Trends Analysis Tool Algorithmic Impact Assessment) and discussed how we have had our own experience litigating a case involving risk flags that were made visible (accidentally) in GCMS, and which may have contributed to a refusal rendered on entirely different grounds than the risk indicators indicated. Through my discussion of bulk decision-making, we discussed how refusals are grouped into buckets that could be informed by the presence of things such as risk indicators.

We know from the IRCC’s response to CIMM Study 8, the following:

  • For indicator management within the Chinook tool, risk indicators are used to notify officers oftrends that IRCC has detected or highlight a particular factor of concern, not to sort visaapplications. Keywords are also used to identify positive considerations such as applications thatmay require expedited processing (e.g. conferences, weddings).

  • Risk indicators are identified and submitted for entry into Chinook by IRCC officers. Indicatorsand keywords are not created by the Chinook tool.

  • The release of specific keywords connected to investigative techniques, trends, and risk profilescould encourage fraud or facilitate the commission of an offence, and are therefore not releasedper section 16(1)(b) of the Access to Information Act.

  • Statistics on the use of indicators and keywords are not tracked globally. If indicators or keywordsare present on an individual application, they would appear in notes in GCMS. Where there are noindicators or word flags on a case processed with Chinook, a ‚ÄúN/A‚ÄĚ (not applicable) would appearin the relevant field in GCMS.

However, because of the function of the information being redacted, in only rare ‘accidental’ disclosure cases have we seen what these look like and so far, without the detail of the actual words or combinations flagged.

As far as I am aware, this is the first case where we have seen the actual directive or recommendation to create risk indicators as a function of data, and in cases where the risk (or adverse outcome) is seen as a student making a refugee claim.

My big question is about the data. I provided one example above where I challenge an alleged causation between refugee claimants and non-compliant studies. I can think of other issues, such as how IRCC collects data on whether someone is Tamil are not (presumably through language), but from my own personal knowledge of agents playing a large role in applications arising from Sri Lanka, are there possible assumptions being made on forms that are not competently filled out? Are there missing disaggregations?

The follow-up question is data-vetting. If these indicators are not being tracked, but simultaneously there are is a six month review period for these indicators, who (if anyone) is in the room to interrogate this data or allege possible bias or problematic collection?

Given the power of these indicators to essentially take applications out of the assembly line to approval, I would suggest much more transparency, and a robust and publicly explainable data review process, needs to be published by IRCC to alleviate concerns that myself and other colleagues have about this process.