Artificial Intelligence (AI) in IoT is a unique aspect of information technology that reacts and functions similarly to a humans? mind, creating automation. AI in IoT enables the emulation of human intelligence to solve problems and learn to understand the high level of activities based on human-driven aspects, decision making, and the emotional cycle. However, users are yet to realize the full potential of AI in IoT systems, rendering it hard to discover and develop effective solutions to the multiple challenges encompassing this emerging technology. The challenges and issues are numerous, including difficulty in data acquisition and storage, ethics and morality issues, inadequate computation speed, flawed algorithms, weak controls and standards to help to cope up with the emerging technologies of AI in IoT, meeting human level, heavy power requirements for computing, and limited knowledge. Fortunately, some approaches can be employed to mitigate the mentioned problems and evade the adverse effects that may result from them. Examples of these approaches include establishing a long-term vision of implementation, engaging different stakeholders to support collection of huge amounts of data, controlling compliance and governance in an organization, and employing software systems to help eradicate flaws in programs.