Search

We are grateful for the opportunity to participate in the Canadian Government’s consultation on a modern copyright framework for AI and the Internet of Things. Below, we present some of our research findings relating to the importance of flexibility in copyright law to permit text and data mining (“TDM”). As the consultation paper recognizes, TDM is a critical element of artificial intelligence. Our research supports the adoption of a specific exception for uses of works in TDM to supplement Canada’s existing general fair dealing exception. Empirical research shows that more publication of citable research takes place in countries with “open” research exceptions -- that is, research exceptions that are open to all uses (e.g. reproduction and communication), to all works, and to all users. Empirical research also shows that text and data mining research is promoted through exceptions that more specifically authorize text and data mining research. While these studies are preliminary and we are still improving on them, they provide evidence that supports the approach of combining a general research exception with a more specific data mining exception.

In crafting new paradigms to balance the often competing demands of technological advancement and privacy, it is critical to treat the issue from an international perspective. Differences in politics, culture and even views regarding the benefits of certain technological advances argue against the creation of a truly universal paradigm. Using the differing treatment of individual privacy and service provider liability as analytical paradigms, this article contends that the lack of international standards at this time might not be so problematic as appears at first glance. Regulation before a particular technology is understood can result in over- or under- protection of privacy rights. Moreover, the present lack of international standards allows for greater experimentation in crafting acceptable boundaries in privacy protection, including, critically, focusing on greater consumer awareness of the intersections between technology and privacy. The international implications of the intersections between technology and privacy will only continue to grow. Part of the solution is to begin to create an international dialogue with a goal to establishing a middle ground and to include multinational organizations as part of the dialogue.

Fair.com? reviews all 3094 ICANN Uniform Domain Name Dispute Resolution Policy (UDRP) decided through July 7, 2001 finding that that the allocation of cases may be unfairly biased toward trademark holders. It arrives at that conclusion based on three key findings. First, forum shopping has become an integral part of the UDRP. ICANN delegates the administration of the UDRP to four dispute resolution providers. Each provider maintains a roster of panelists who serve as judges in deciding the disputes. Complainants, who are invariably trademark holders, have the right to select which provider will handle their case. Track records of the various providers are a matter of public record and show that 90% of complainants rationally choose the two providers who feature panelists that render the most pro-complainant decisions. Second, there is a correlation between the selection of panelists and case outcome. UDRP cases are decided by either single or three-member panels. Single member panels feature one judge chosen exclusively by the arbitration provider. Three-member panels contain three judges selected by both the complainant and the respondent. This is significant since data shows that when providers control who decides a case, complainants win just over 83 percent of the time. As provider influence over panelists diminishes, such as in three-member panel cases, the complainant winning percentage drops to 60 percent. In an attempt to explain this, the study arrives at the third key finding - the higher winning percentage in single member panel cases may stem from provider bias toward ensuring that pro-complainant panelists decide the majority of cases. Although the allocation of cases is supposed to be random, the study finds that this may not be the case. For example, the National Arbitration Forum, one of the larger dispute resolution providers, has 135 panelists. However, 53 percent of its single panel cases are decided by only six of them. Those six panelists rule in favour of complainants 94 percent of the time. The article concludes that the solution to the forum shopping issue, and with it the concerns about bias and inconsistency within the UDRP, is surprisingly simple -- all contested UDRP actions should involve three-member panels. Establishing the three-member panel as the default would remove most provider influence over panelist selection and ensure better quality decisions by forcing panelists to justify their reasoning to their colleagues on the panel. As with the current system, both parties would play a role in selecting one panelist, who may be part of any ICANN-accredited providers' roster, while the provider would select the third panelist from among a list that both parties have reviewed and accepted.