Thank you very much, and good morning.
As you heard, my name is Michael Geist. I'm a law professor at the University of Ottawa where I hold the Canada research chair in Internet and e-commerce law. My areas of specialty include digital policy, intellectual property, and privacy. I have appeared before this committee many times on IP issues and, as always, I appear in a personal capacity representing only my own views.
I'd like to start by welcoming this committee's study on an important aspect of intellectual property. However, I respectfully suggest that the name of the study gets it wrong. I understand the notion that tech transfer has taken hold in some discussions on how Canada can shift innovative research from Canadian campuses to exciting new commercialization opportunities. I'd like to suggest that the real goal is not tech transfer, but knowledge transfer.
Knowledge transfer encompasses a far broader set of policy goals that seek to take the knowledge that emerges from within our labs and classrooms and bring it out to the public, whether for commercialization, better public policies, or a more informed and engaged public. Knowledge transfer certainly includes tech transfer, but it also includes research papers, data trials, educational materials, and highly qualified students and personnel. Simply put, if the target is just IP and tech transfer, we miss out on many of the benefits that come from innovative post-secondary research and run the risk of establishing the wrong incentives within our policy framework.
Further, the potential emphasis on the U.S. Bayh-Dole approach, in my view, is misplaced. As you heard from department officials earlier, there is little evidence that the policies governing who owns IP rights have an overriding impact on the success of tech transfer as measured by volumes of patents and licences.
This should come as little surprise to anyone who has spent time on campuses with academic researchers. The metrics of success in the academic environment, such as publications, grants, tenure, chairs, and successful students, have little correlation with commercialization. Even for those with commercial interests, those are often achieved through consulting arrangements or other mechanisms where the business expertise is by and large left to business people.
I would argue that the emphasis on university-based patenting is misplaced. It can have a corrosive effect on universities, that forgo important publicly funded research in favour of potential licensing or patenting opportunities. With properly funded institutions, there is no need to chase licensing dollars. Instead, cutting-edge research ends up in the hands of businesses that can better leverage it for commercialization opportunities. This should not be viewed as lost revenue for universities or their researchers, but rather as a better return on the public's investment in post-secondary research.
From an IP strategy perspective, I'd like to focus on two broad issues. The first is open access publishing. If the currency of academics is publishing, not patents, then the challenge is how to ensure that the published research ends up as broadly distributed as possible. While it has captured limited attention outside educational circles, the Internet has facilitated the emergence of open access publishing of research, transforming the multi-billion dollar academic publishing industry and making millions of articles freely accessible to a global audience. The move toward open access means that global research is far more accessible to everyone: scientists, researchers, businesses, and the general public.
The three federal research-granting institutions, CIHR, NSERC, and SSHRC, have adopted open access mandates that require recipients of federal funding to make their published work available under open access. This helps foster greater collaboration between researchers and the business community, with improved access leading to commercialization opportunities that might otherwise be missed.
Further, openly available articles are already being incorporated into teaching materials, thereby replacing conventional textbooks and removing the need for copyright permissions and fees. As for government strategies, open access mandates should only be the beginning. Moving toward open trial data and open book publishing are the next steps in linking significant public funding to enhancing public access to their investment.
From an intellectual property legal barriers perspective, we should start by noting that Canada already meets or exceeds international standards on IP. A key concern, if we're looking to address it, is the abuse of IP rights that may inhibit innovation. The Canadian government could address the issue through an anti-IP abuse law.
There is no shortage of policy possibilities. We can talk about all three areas, but just to give patents as an example, countering patent trolls could include a prohibition against legal demands that are intentionally ambiguous or designed to induce a settlement without considering the merits of the claim. Other reforms could include requiring public disclosure of demand letters, reforming the Competition Act to give the Competition Bureau the power to target anti-competitive activity by patent trolls, and giving courts the power to issue injunctions to stop patent trolls from forum shopping.
There is also a need to address IP barriers that may limit the ability to take research from labs into the commercial world. For example, the federal government this year placed a big bet on becoming a world leader in artificial intelligence, AI, yet restrictive copyright rules may hamper the ability of companies and researchers to ultimately test and bring new AI services to market.
What does copyright have to do with AI? Making machines smart, whether engaging in automated translation, big data analytics, or new search capabilities, is dependent upon the data being fed into the system. Machines learn by scanning, reading, listening, or viewing human-created works. The better the input, the better the output. Copyright law crops up because restrictive rules may limit the datasets that can be used for machine learning purposes, resulting in fewer pictures to scan, videos to watch, or text to analyze.
Given the absence of a clear rule to permit machine learning in Canada, often called a text or data mining exception in copyright law, our legal framework trails behind other countries that have reduced the risks associated with using datasets in AI activities.
There are two ways to overcome the copyright AI barrier.
First, Canada could, and I would argue should, emulate the U.S. fair use model by making the current list of fair dealing purposes illustrative rather than exhaustive. The U.S. exception is open to any purpose, as striking a fair balance depends upon the use of the work, not the purpose of the copying. Since machine learning does not harm the primary purposes of the original work, most text and data mining will qualify as fair use.
Second, other countries have tried to address the issue by creating a specific exception for text and data mining or computer informational analysis. For example, Britain's exception allows copies of works to be made without permission of the copyright owner for the purposes of automated analytical techniques to analyze text and data for patterns, trends, and other information. We don't have a similar provision under Canadian copyright law, and as we seek to move from the lab into commercialization opportunities, that may inhibit the ability to do so.
I look forward to your questions.