As artificial intelligence (AI) continues to make inroads, data have taken on a crucial economic role in many businesses, including those active in the aerospace industry. Who owns these data and the rights associated therewith has become an inescapable question.

In most jurisdictions, including Canada, the legislative framework has failed to keep pace with technological progress. There is no data appropriation mechanism just yet. And the limits set on today’s intellectual property (IP) frameworks are such that the scope of data rights remains uncertain.[1]

Copyrights, which protect all “original” literary, artistic, dramatic or musical works, have proven to be inadequate when it comes to AI. In fact, most jurisdictions will only attribute copyrights where there is a human author. If there is no human intervention in the creative process, the results generated by an automated IA process might not qualify for copyright protection. What is more, even if data compilations do qualify for copyright protection, the data making up the compilation do not. And finally, the standard of originality might not be satisfied in certain contexts, notably when data originating from similar devices are compiled, or when the data are arranged according to specific industry norms or standards.

Another possible option appears to be trade secrets. This form of IP, however, protects the data’s confidentiality and not the data themselves. In other words, once confidentiality is lost, no ownership is attributed to the data. If the data’s secrecy is preserved, it can be difficult in some industrial environments to use these data to develop products and services. And even if the use of a trade secret is part of a company’s business model, the measures required to ensure and prove secrecy are often onerous, even prohibitive.

Given the limits of today’s legal regimes, legislative evolution is in order. But while we wait for this transformation, the best way to ensure legal security is to provide for it by contract. To control and protect data, the various stakeholders will need to agree on contractual provisions that clearly specify data ownership and what constitutes authorized data use. In a multi-stakeholder ecosystem where every intervener can claim ownership rights, it will also be important to identify any person who contributed to the technology, while taking into account that technology’s ability to evolve.

Author : Awa Carole Diop

[1] “Data Ownership”, Teresa Scassa, Center for International Governance Innovation (CIGI) Papers No. 187, September 2018