Investing in your Firm's AI Future

When it comes to the business aspects of AI, investments come with substantial risks and uncertainties, and managers do need careful evaluations with these investments. Current understandings of how managers’ perceptions of AI influence their investment decisions are limited due to the rapidly increasing changes of the industry.

Firms are investing heavily in AI technologies, including robotic technologies and machine learning automation, to perform cognitive tasks similar to people. With the increasing sophistication of AI, more specifically organizational AI, managers are still struggling to assess its potential for firm value creation, especially given AI’s ability to significantly improve cost efficiencies and decision-making processes.

 In “Manager Appraisal of Artificial Intelligence Investments,” Walton College Assistant Professor of Information Systems Abhijith Anand and his coauthors Magno Queiroz (Florida Atlantic University) and Aaron Baird (Georgia State University) develop a classification system for AI based on action autonomy and learning autonomy and propose how managers’ delegation preferences influence their AI investment appraisal.  

Research to date has focused on the post-adoption stages of investments, whereas Anand and his coauthors analyze the pre-adoption phase. By shifting the focus from the outcomes to the initial allocation of AI investments, Anand and his coauthors offer new insights into the appraisal process and set an agenda for future research on AI investment.

 To better understand how managers’ decisions before adoption impacts their firms, the authors focus on how managers’ AI preferences and value creation priorities influence their investment. 

Challenges and Opportunities of AI

With rapid advancements in machine learning and the rise of big data in the past decade, AI systems have evolved to mimic human cognitive functions, allowing companies to innovate and adapt by augmenting or automating tasks traditionally performed by humans. Unlike traditional information systems, AI operates with higher autonomy, which means managers must weigh the trade-offs between the loss of control and potential benefits AI can bring.

For example, Netflix uses AI for content production and advertising, IKEA and Wayfair use it to create interior design services, and Siemens and Pfizer use it in healthcare for diagnostics and drug development.

 Managers need to evaluate the full range of AI capabilities, that could range from simple personalized pricing decisions to complex data mining processes for opportunity recognition,all while also considering uncertainties related to AI’s accountability and accuracy. AI investments can enhance the routines of employee processes, freeing up managers for more creative tasks. However, they also pose risks in non-systematic cognitive tasks, such as self-regulation, optimism, hallucination, and tenacity – areas where AI still falls short.

When considering AI adoption, managers must also evaluate how it affects power dynamics and accountability within their organizations. For instance, if an AI system makes decisions or takes actions that result in harm, managers must evaluate responsibilities of the employees who relied on it for key decisions. Therefore, to maximize business value, it is important to understand the reasons behind managers' AI investment decisions while ensuring that human agency is factored into AI investment appraisals.

 The study suggests that the appraisal process is shaped by managers' individual perspectives, particularly in how they value past experiences, present demands, and future opportunities. Managers’ focus on time influences their approach in three ways: iterative (past-focused, relying on established protocols), practical-evaluative (present-focused, addressing current urgencies) and projective (future-focused, embracing experimentation). The managers' dominant viewpoint affects their willingness to pursue these AI investments after reviewing challenges and opportunities of the investment. Managers who are future focused tend to take more risks regarding AI agency, whereas those who focus more on current pressures or past experiences with emerging technology tend to err on the side of predictability.

AI and Business Investment

AI can drive efficiencies in the relationships of talent acquisition, development of drugs, demand forecasting, inventory management, auditing and performance management. Externally in the market, AI can enhance price adjustments, introduce new services, tailored product offerings, and marketing campaigns, leading to faster market response, increased customer satisfaction, and sustainable competitive advantages.

Managers should undoubtedly strategically allocate resources to AI investments that are most likely to generate business value. The resources are allocated without specific AI investments identified, requiring managers to justify AI investments based on the potential value. This could lead to high risks if there is failure in the AI projects, particularly when managers invest in AI to enable non-systematic cognitive tasks. For example, a major cancer research center invested $62 million to create an effective AI-based cancer diagnosis system only to have their efforts prove unsuccessful after four years.

Additionally, managers must choose between using AI to automate tasks or augment human capabilities. This choice heavily impacts their investment decisions regarding what would be best for the business. These decisions are crucial not just for individual firms but for entire industries as AI shifts the control and responsibility from managers to digital systems, raising wide concerns about liability and adaptation.

Furthermore, managers’ preferences for delegating tasks to AI influence their investment appraisals. Anand and his coauthors helpfully introduce a classification system for self-learning AI. They also propose that managers’ appraisal of AI investments is not only shaped by the type of business value they aim to create but also whether managers are past-, present-, or future-focused.

Future Directions and Considerations

What makes Anand and his coauthors’ research impactful for firms is the focus on the decisions and preferences before an AI solution is adopted. But they also call out several opportunities to further their research by examining managers’ temporal tones using field surveys, archival data, or through case studies exploring retrospective assessments of AI investment decisions to determine whether actual AI investments align with dominant temporal tones. Another area of significant opportunity is office politics, as the decisions made by managers are often made on behalf of the other managers.

For firms, Anand’s research is valuable because it explores the largely uncharted waters of pre-adoption AI decisions. The hype cycle affects everyone, both managers as well as investors, and both should be aware of the possible pros and cons of adopting the use of AI investments in their business practices. Instead of focusing solely on implementation or integration with existing solutions, firms should also think about the questions they are asking (or not asking) and assumptions they are making before adopting an AI solution. By developing a more informed outlook on AI appraisal, their research urges managers to shift toward the possibilities of what AI could do for their competitive advantages in today's world.