The present study undertakes the analysis of whether generative AI (GenAI) can be applied to improve organizational productivity and outputs. Although the results of the quantitative analysis performed in six organizations prove that the task efficiency improvement rates are high enough to mention (15-40 percent), especially in customer service (25-40 percent faster response, 10-15 points rise in CSAT) and content creating (30 percent faster), the research exposes socio-technical and governance gaps that are likely to compromise the long-term success. There is a sharp disjuncture between the emphasis on automation amongst executives (83%) and reality in operations: processes that were fully automated (e.g., chatbots) had 22 percent more error/ rework rate compared to augmented human expertise utilizing GenAI in preparatory work (e.g., report drafting) which increased high-value work by 35 percent. Also, 67 percent of operational users experienced biased results (gender, cultural) resulting in customer complaints (50 percent) and delays in resolutions (35 percent), which was compounded by the lack of bias mitigation and data auditing. Ethical governance was performative ("ethical theatre"); thus, a policy, inability to quantify different KPIs, training (78 percent of staff were untrained), and management led to a ease of compliance infringements (e.g., 120 HIPAA breaches). More importantly, organizations undertaking reskilling programs had achieved 45% higher adoption and 30 percent more innovation whereas organizations that had cross-functional ethics committee achieved 40 percent fewer compliance incidences. Results show that achieving the productivity potential of GenAI through the transition to human-AI augmentation rather than workforce reduction, and significant investment in reskilling, continual bias-mitigation, and ethical governance with accountability, are necessary to address the risks to GenAI-related productivity. The application of technology is not the only prerequisite to attainment of sustainable gains, but also improvement in the area of human infrastructure known shortcomings.
Mariam Al Kuwaiti (Sat,) studied this question.