Fairness or avoidance of bias means that the AI system is sufficiently statistically accurate and avoids discrimination and that it considers the impact of individuals’ reasonable expectations. Also, this can imply that AI shall be in compliance with Law against discrimination, or in other words Law against discrimination should be amended to comprise AI. Last, but not least, AI systems should be safe and secure, in accordance with technical standards and to ensure the protection of human rights in digital world. Also, in this line there is a requirement for human control over AI systems, meaning that AI should be permitted only to recommend course of action for human, but never to take actions without human impute or instead of human impute. 38 This also means that developers of this systems should ensure that humans should be able at any time to suspend and amend any operation of AI. VII. CONCLUSION This paper analysed potential that AI has for the development of Montenegro, ecological country dedicated to sustainable development. By enhancing productivity and administration, AI technologies can greatly advance the travel industry. However, the application of AI may give rise to grave ethical and legal issues regarding algorithmic bias, fairness, accountability, transparency, and data privacy. Despite AI is being recognized in many strategic documents, Montenegro still lacks legal framework for the use on AI, which can cause legal uncertainty and serious threats to human rights in Montenegro. It is concluded that Montenegro should harmonize its laws with EU standards, in particular with GDPR and EU AI Act. This can further enhance confidence of foreign investors, significantly support tourist experience and promote Montenegro as a destination for smart tourism. 38 Matt Hervey and Matthew Lavy, The Law of Artificial Intelligence, Sweet&Maxwell, p.107. 106 MILANA RANKOVIĆ | JAKŠA VUKOVIĆ
RkJQdWJsaXNoZXIy MTE4NzM5Nw==