Today’s hiring landscape presents organisations with increasingly complex and sometimes paradoxical challenges. On one hand, there is a growing imperative to embed diversity, equity, and inclusion (DEI) into every part of the talent identification process. This includes developing assessments that enable candidates of all neurotypes to demonstrate their abilities fairly. On the other hand, the rapid rise of generative AI (GenAI) tools such as ChatGPT has created new concerns about the integrity of unsupervised, online assessments and the authenticity of candidate responses.

As organisations strive for greater neurodivergent inclusion, selection methods must evolve to accommodate diverse cognitive styles. Traditional tools like situational judgement tests (SJTs), often built around neurotypical assumptions, can unintentionally disadvantage neurodivergent candidates (McMillan et al., 2023). Simultaneously, GenAI tools make it easier for candidates to generate or enhance responses in unsupervised contexts, raising concerns about assessment security and validity (Harwood et al., 2024). In 2026, talent professionals are required to navigate both inclusion challenges and technological disruption.

SJTs remain widely used due to strong face validity and candidate acceptance (McDaniel et al., 2007), yet their reliance on text heavy scenarios and predictable rank order formats makes them increasingly vulnerable to GenAI assisted responses (Mittelstädt, 2024). These same design features can also create cognitive overload or disadvantage candidates with differences in executive functioning, information processing, or social cognition (McMillan et al., 2023). While these issues are often treated separately, they are interconnected, prompting the central question: how can SJTs be redesigned to be both more neuro-inclusive and more resistant to GenAI, while maintaining psychometric rigor?

Our case study explored redesigned SJT formats across content presentation, response modality, and scenario design. A study with 557 participants compared traditional ranking SJTs with the adapted format. Findings suggest the new format maintains comparable psychometric performance while improving resistance to casual GenAI use and reducing cognitive load. Initial findings suggest that participants preferred the redesigned items, describing them as clearer and less mentally taxing. These results demonstrate how thoughtful assessment innovation can enhance both fairness and security without compromising validity.

The rise of GenAI, alongside increasing demands for inclusion, requires a new mindset where inclusive design can serve as a defence strategy. Participants will leave the session with concrete questions to help them identify gaps and opportunities in the current landscape of talent assessment strategy, and a clearer understanding of what future-ready assessments look like in practice.