Artificially intelligent investment advisers which are commonly called robo-advisers are Artificial Intelligence or Machine Learning-based online platforms that provide investment advisory services to investors. Both in the USA and Canada, robo-advisers are required to adhere to the regulations governing investment advisers. In the USA, the Securities and Exchange Commission has clearly stated that robo-advisers owe a fiduciary duty to their clients. In Canada, while there is no statutory fiduciary duty, according to the jurisprudence, based on the facts and circumstances of an adviser-client relationship, an investment adviser may be required to act as a fiduciary. This Thesis argues that Canadian robo-advisers, too, owe a fiduciary duty to their clients. Given that the securities regulations and the standards of conduct under the investment adviser fiduciary duty have been developed throughout the history with an eye to how humans behave and what kind of motivations they have, the use of algorithmic and AI-based systems as the provider of the investment recommendations raises concerns as to these state-of-the-art financial intermediaries' capability of complying with the securities regulations and acting as good fiduciaries. This Thesis conducts an analysis of robo-advisers under the US and Canadian securities laws to illustrate why robo-advisers are/should be required to act as fiduciaries in the USA and Canada, identify and analyze the risks and challenges that may negatively affect a robo-adviser's compliance with laws, and explore implementable regulatory solutions to address these risks and challenges. This Thesis argues that there exists a number of issues that may adversely affect a robo-adviser's compliance with the regulatory requirements as a good fiduciary, and recommends the intervention of the US and Canadian securities regulations into robo-advisory in a number of ways and methods.