The COVID-19 pandemic accelerated the shift to online assessment, prompting debates over validity, security, and increasingly the impact of Artificial Intelligence (AI) tools, especially generative AI, on traditional examination methods. This paper explores perceptions of the evolving landscape of online assessment and the role of AI within higher education, building on work conducted at the University of London and the Open University UK. Workshops used speculative methods to envision potential future scenarios and gather perspectives. These revealed a complex, ambivalent outlook on online assessment and AI's role in education. The paper highlights the polarised views surrounding AI, ranging from ethical concerns about academic integrity and unfair advantages to opportunities for enhancing learning and inclusivity in assessment practices. Our findings reflect attitudes of students and educators towards AI and online assessment, identifying key themes such as ethics and integrity, the need for redesigning assessments, issues of diversity and inclusion, and the dependencies required for successful integration of AI. Participants highlighted both the potential benefits of AI in creating more authentic and personalised assessment experiences and the risks of exacerbating inequalities and undermining institutional credibility. The paper underscores the urgent need for a balanced approach to AI and online assessment in educational policy and practice, emphasising inclusive, ethical, and innovative approaches to navigate the challenges and opportunities presented by these technological disruptors.