Policy, Progress, and Protection
Technological security comes from being the first mover
In 2012, I took part in a protest against the closure of the Experimental Lakes by the Harper government, an attempt to protect a globally unique scientific resource. This was a critical moment in the push back against long-standing muzzling of government scientists by a government that wanted to control parts of the scientific narrative. The protest was unsuccessful, and the Experimental Lakes were closed. While they were ultimately reopened 2 years later and continue to operate today, it is arguable that the damage was done. At best, Canadian environmental science was set back years.
When the Trudeau government took over, part of their platform included a commitment to ensuring that government scientists would not be muzzled by political concerns. Part of the policy response intended to improve the situation and rebuild trust in government as a contributor to science was the NRCan Policy on Scientific Integrity. According to the stated goals, this was set out to ensure that scientific research could be conducted by NRCan scientists without political interference, captured primarily by the following two statements:
6.2. Ensure that the conduct of NRCan research and science and any research or scientific products, as well as any associated communications, are free from political, commercial, client and stakeholder interference.
6.3. Ensure NRCan research and scientific information is made publically available in a timely manner and in keeping with the Government of Canada's Directive on Open Government, unless there are clear and compelling reasons for limiting disclosure.
I was recently approached by a government employee, speaking on the condition that they not be named, who drew my attention to a more recent piece of departmental policy, the NRCan Policy on Management of Intellectual Property Resources, the most recent iteration of which was made effective April 2023. This is a piece of policy which lays out how intellectual property arising from government-conducted science is to be managed, and it draws on the expectations laid out in the NRCAN Policy on Scientific Research above, as well as a host of other policy frameworks, including, among others, the Open Science Action Plan, the Policy on Government Security, the Procedure for Publishing, and the Policy on Title to Intellectual Property Arising Under Crown Procurement Contracts.
This last policy is notable in that section 2.5 specifically states that collaborative research agreements are out of scope. In the NRCan Policy on Management of Intellectual Property Resources, this distinction is lost. Whereas normally a child policy should have a scope of application no broader than the parent, the loss of these distinctions broadens the scope of application of the departmental policy to cover collaborative research projects.
According to my source, it has repeatedly occurred that research plans submitted for funding have been bogged down in red tape using this policy. Their interpretation, given the expanded scope of the policy causing the delay, is that this is happening in cases involving collaborators deemed unsuitable or research that is too exploratory to clearly define IP outputs, but minimal feedback has made addressing the specific concerns a challenge. This results in effectively stalling projects, while technically not violating 6.2 and 6.3 of the Scientific Integrity Policy noted above. This shift has coincided with a centralization of decision-making power and more vertical, top-down control over the research process. Government scientists often have multiple layers of management above them, not all of whom are scientists themselves and who often require higher security clearances, leading to difficulties communicating and slowdown of processes.
“What appears to be happening now is that it seems like we’re heading toward wartime, and wartime policies”, they observed. “Many of the things that they have activated, that have been activating for the last year or so, have been of this type.
At the very least, these outcomes are incompatible with the spirit of the NRCan Policy on Scientific Integrity, though my source indicated that this policy is currently being revised. I will be watching this process closely going forward.
Technological Security and Academic Freedom
While control over academic research in normal circumstances is very light touch and administered mainly through funding body decisions, several recent policy changes are taking a much more direct, top-down hand in controlling Canadian research. Notable among these is the establishment of the list of Named Research Organizations that are deemed to be a security threat and which are ineligible as collaborative partners, and recent decisions relating to international student participation in Canadian academic research. This is consistent with a need for security in the face of a future with significant short-term uncertainty, and in an era where technological advancement dramatically outpaces policy response, it is understandable that the policy side wants to pump the brakes.
I do not have access to the same level of information as those who make these decisions and so I am not in a position to provide an opinion on the need for greater security in relation to dissemination and sharing of Canadian research outputs. However, I believe that prevention of forward progress at the level of basic science is counterproductive regardless.
I am generally of the belief that development of any technology that is sufficiently desirable and has funding behind it is just a matter of time. Whether it is AGI, room-temperature superconductivity, or fusion, as long as it is physically possible, has value, and scientists somewhere have enough funding to move the needle, it will be developed sooner or later. (Don’t confuse this belief with techno-optimism. I do not necessarily believe this is a good thing, simply that it is).
The necessary first step to securing and controlling a technology is to understand it. This provides a first-mover advantage in terms of securing novel technologies, as the first player to develop something can use their head start to build thoughtful policies and safeguards around it while others are still in the development phase. While blocking development or contribution to development of a technology domestically may slow its arrival, it eliminates the possibility of evidence-based, proactive policy development to manage the results. As is becoming increasingly clear as AI advances in leaps and bounds, this is not an option that is compatible with technological security.
Even in cases where security concerns impose reasonable delays on the dissemination of research results or exclude specific groups from collaborating in the development process, it is important that this not interfere with domestic scientific advancement. The reasoning is very simple: Canadian policy that hinders scientific advancement cannot prevent forward progress abroad. All it can do is prevent Canada from keeping up.
Technological security and academic freedom are not mutually exclusive. There are productive ways to secure research and policies that allow for delays in public dissemination of research outputs in the name of security, but preventing forward progress entirely is not among them. Canada may have already missed its opportunity to establish a leading position in AI policy development, and we are already behind the rest of the OECD on every metric of innovation that matters. Let’s not institutionalize practices that reinforce these issues.
If you can shed more light on these issues, I want to hear from you. Please reach out to me directly here.