Establishing Trust in the Future of Enterprise AI

In recent years, we have seen a rapid increase in the development and implementation of artificial intelligence (AI) in various industries. With groundbreaking events like the AI Security Summit held at Bletchley Park last week, many of us seem eager to join the digital revolution. However, like any new technology, there is still a certain level of skepticism surrounding the advancements in AI. Some businesses hesitate to adopt AI on a large scale due to concerns about its reliability, transparency, and fairness in a production environment. Rajan Padmanabhan, Head of Industry Architecture, Data, and Analytics at Infosys, explained in an interview with AI Magazine why some organizations seem reluctant to embrace AI even in this digital age. "What are the concerns that users have about AI technology?" Padmanabhan explained that there is a perception that AI is designed to replace humans. "Research shows that automation will lead to redundancy in certain jobs, exacerbating anxiety about the emergence of AI. However, skeptics often overlook the role of AI in enhancing human intelligence and improving efficiency in repetitive tasks. The development and deployment of AI technologies will create new jobs that do not exist today." In addition, ethical issues and risks associated with AI deployment have also raised concerns. "Several AI pilots have discovered flaws such as cognitive biases (e.g., racial recognition) embedded in algorithms and models," said Padmanabhan. "These have led to unforeseen consequences and raised ethical questions around using AI as a reliable decision-making tool. Businesses can overcome these biases by investing in machine learning frameworks and risk-based approaches to develop AI tools." As with any new technology, concerns about personal data privacy naturally arise, especially with the increased complexity of cybercriminals due to AI. Padmanabhan, who leads architecture, solution strategy, and innovation centered around data, AI, and cloud, believes that "given the primacy of data in AI, access and use of personal information have always been a contentious issue. Additionally, regulatory bodies like the European Union have introduced strict laws to protect personal privacy, such as the General Data Protection Regulation (GDPR). These factors directly impact the adoption of enterprise AI." "What can businesses do to overcome skepticism surrounding AI?" To overcome the skepticism surrounding AI adoption and demonstrate its value, Padmanabhan explained that pilot projects and case studies could be part of the solution. He said, "According to a survey by the European Commission, 56% of businesses deploying AI technologies in Europe plan to scale up in the next two years, while only 4% plan to scale down. This is a significant figure as it indicates the proportion of businesses that have overcome inertia and skepticism. Such pilots and demonstrations enable business leaders to present strong cases that go beyond concepts and anecdotal evidence, entering the realm of evidence and business value, which is crucial for the adoption and successful deployment of AI." Padmanabhan also believes that successful AI deployment requires several elements to remain consistent and synchronized. He said, "Businesses must develop custom frameworks that suit their specific needs and purposes. This issue must be addressed from two perspectives. First, adopting AI among employees requires them to embrace a data-driven culture and enhance their skills. Second, democratizing technology within the organization makes it easier for employees to access technology, empowering them and decentralizing decision-making." "AI is not just about technology; it is also about being human-centric," Padmanabhan emphasized. "One of the biggest challenges AI faces in the future is the perception that it is a zero-sum game with only one winner – either humans or machines. As research and recent developments have shown, this is far from reality. Most conversations surrounding AI revolve around data and emerging technologies. This perspective overlooks or neglects the human aspects, which are equally important for any organizational scope and transformative intervention, such as enterprise AI. For example, gaining acceptance of AI requires building trust through transparency and gaining stakeholders' recognition of acceptable AI use for different purposes." In conclusion, Padmanabhan summarized, "Just like most groundbreaking technologies developed throughout history, what matters is not the innovation itself, but how we apply it. AI is no exception. Ultimately, effective management of AI by businesses will determine how much of the $14 trillion total value they can capture."