Like many industries, financial services continues to face recruitment and retention challenges.1 Although artificial intelligence may be able to help address some of these issues, it remains relatively underutilized in this area. A recent Deloitte survey revealed that only 18% of executives in the financial services industry said their organizations are implementing generative AI within the talent function, compared with 47% that are using gen AI in marketing, sales, and customer service.2
To learn more about current and planned usage of AI in talent management, the Deloitte Center for Financial Services conducted an online focus group with 38 senior executives from some of the largest financial services institutions in the United States (see methodology). While most participants were talent leaders and the rest served in IT or strategy leadership roles, all had visibility into or decision-making authority over how AI could be used to enhance the talent experience.
Overall, many participants believe their organizations are just scratching the surface when it comes to applying AI to enhance the talent function. They highlighted the current state of deployment and areas where AI could potentially add more value further along in the talent lifecycle (figure 1). Participants believe AI could be leveraged to significantly boost employee engagement and retention, pointing to performance evaluation, succession planning, and employee flight risk as primary focus areas.
Many participants told us that their organizations are currently leveraging AI predominantly to help attract and recruit talent. They report that deploying AI in this area has yielded a strong return on investment; it has improved skills assessment and capturing competencies better than using traditional methods. “I have seen the most significant return on investment in the areas of talent acquisition and learning and development,” a participant who serves as head of business, HR/talent management at an investment management firm said. “For example, by implementing AI-driven recruitment tools, we have reduced our time-to-hire by 18%.”
Gen AI is enabling a shift from pull to push people analytics. Historically, talent executives have had to pull information, analyze it, and derive meaningful insights. Now, with just a few prompts, an algorithm can monitor vast amounts of data, identify anomalies, highlight points that may be of interest, and generate reports.3 Having more access to real-time and increasingly nuanced insights can help talent executives proactively collaborate with business functions and executive leadership to help mitigate preventable employee turnover.
Participants agreed that AI could be a valuable tool to help retain employees. They mentioned that being able to identify early indicators of disengagement and burnout, a significant challenge for organizations, would be particularly helpful. Often, organizations invest heavily in wellbeing and other similar programs without fully understanding the key factors driving burnout and resignations.4 Identifying these root causes could enhance the effectiveness of these initiatives (figure 2).
Talent executives can use AI to analyze a multitude of data points, including:
By overlaying these data points with demographics, such as age and education, and external inputs, such as labor market trends and macroeconomic conditions, talent leaders can view a more comprehensive picture of employee retention or flight risk. A participant who serves as a director of HR/talent management at a bank explains: “Gen AI can predict the likelihood of an employee leaving by identifying predictive factors such as employee burnout, declining performance, decreasing engagement. Also, a predictive model could assign a flight risk score to flag employees that are most at risk [of leaving].” Those with a higher score may warrant proactive retention strategies.
Most participants agreed that performance management was another area that could benefit significantly from AI’s capabilities. They highlighted that AI tools can analyze historical data of employees with similar career trajectories and identify common factors correlated with high performance. A CIO from an investment management firm said, “AI can be used to identify high performers and identify key performance attributes and characteristics of those performers and then apply those attributes to other performers for advancement.”
Using technology to uncover key indicators of high performance has worked before. Nearly 15 years ago, Google launched Project Oxygen to find out what attributes differentiate great managers from those who are less skilled. They mined data from employee surveys, performance reviews, and double-blind interviews to identify eight key behaviors shared by high-performing managers and incorporated them into their management development programs. This led to improvement in managerial effectiveness and performance. Over two years, the median employee feedback survey scores increased by 5 percentage points, with the greatest improvements seen among the lowest-scoring managers.6 Today’s technologies can enable initiatives like Project Oxygen at scale.
Furthermore, participants in our focus group highlighted the importance of mitigating bias and ensuring equitable growth opportunities for all employees. They identified two stages in the talent lifecycle—performance evaluations and career advancement—as having the most potential for bias after an employee joins the organization. A participant who serves as VP for HR/talent management from a bank said, “Often when performance is not metrics-based and requires soft skills, or if the skills do not match that of the manager, it can be misjudged. Given gender, age, and neurodivergent and cultural differences, this can create bias and [lead to] unfair, poor ratings of just ‘different’ skill sets.”
Participants were eager to implement AI tools to facilitate a more objective evaluation of skills, experience, and capabilities and minimize unconscious biases. There are some examples of such tools being used for career development. For instance, Singapore-based DBS Bank’s iGrow platform leverages AI and machine learning to help employees identify the skills they will need to achieve their career goals; it also recommends relevant education, exposure, and experience opportunities.7 Further, some widely available tools can proactively detect biased language in performance reviews and suggest more actionable feedback to make performance reviews more effective.8
AI can also be leveraged to help people find mentorship and networking opportunities, which could help boost engagement and support career growth. By using contextual understanding, gen AI-driven mentorship programs can objectively recommend professional networks and sponsorships to help professionals make meaningful connections with senior leaders or sponsors. Mastercard, for example, uses Unlocked, its internal talent marketplace, to match mentors and mentees based on skills and experience. Employees can also explore short-term projects, volunteering opportunities, and open roles through the platform.9
Succession planning—identifying top talent and preparing a comprehensive action plan for top-level transitions—is an important part of any organization’s long-term strategy. But most participants said their organizations still rely on manual methods, which are often time-consuming and can potentially overlook qualified candidates, to handle these responsibilities. Since many organizations are moving toward a more skills-based leadership approach,10 succession planning processes that focus too narrowly on linear and hierarchical career paths may overlook potential candidates who could be considered for upcoming leadership opportunities based on their skill sets.
“Succession planning is still done quite manually, which leads to under recognition of potential internal talent,” a VP of HR/talent management from a wealth management firm said. “Applying skill-based criteria should expand the internal pool beyond the traditional selection criteria (that is, who you know).” Talent leaders can leverage AI to help transform annual succession planning into a dynamic and flexible process that cultivates a strong and diverse leadership pipeline. Several participants mentioned that by utilizing AI’s analytical and predictive capabilities organizations can identify leadership gaps, expand the internal candidate pool, and suggest career paths and training programs to prepare future leaders. The technology can also help talent leaders identify high-potential candidates by providing a comprehensive picture of an individual’s leadership capabilities based on employee feedback, engagement, goals and objectives, and key competencies.
AI can empower talent leaders and boost employee engagement. But it can require a closer analysis of employee behavior, which some employees may find unsettling and invasive. A participant who serves as VP of HR/talent management at a bank said, “The biggest challenge I see would be getting workers to trust the [AI-based] tool and the process and getting them to understand how the tool works.”
How can organizations gain trust without being overly intrusive? How can they ensure the technology is used responsibly and ethically, and communicate this effectively?
Practice proactive transparency: In Deloitte’s 2024 Global Human Capital Trends research, 86% of workers surveyed stated that greater organizational transparency leads to higher workforce trust.11 Organizations should be forthcoming with employees about how and why their data is being used, as well as how the data will be collected and safeguarded. By respecting employee privacy and using the data only for the intended purpose, they can build trust.
Organizations should clearly communicate that the intention behind data collection is to create mutual benefits for both the organization and its employees. Employees may feel uncomfortable with AI tracking and reporting their performance, but if it is used to help them personally understand their performance and provide comparative data without sharing it with others, they would likely find it helpful. For instance, when collecting data for Project Oxygen, Google leadership emphasized confidentiality and issued frequent reminders that the surveys were strictly for self-improvement, not as a performance metric.12
Ensure ethical and responsible use of data: Organizations should ensure proper data management and identify and mitigate real-world biases that may creep into their algorithms and implement necessary checks and controls. Additionally, cross-functional collaboration involving executive leadership, IT, legal, and talent functions can improve governance by leveraging their collective expertise. Equally important, organizations should protect themselves by establishing mechanisms to address potential issues, particularly as regulatory oversight on the use of AI tools in the workplace becomes more stringent.13
Maintain human oversight: Despite the increasing sophistication of AI models, algorithms may still struggle to grasp the nuances of individual growth and development trajectories. This underscores the importance of maintaining human oversight as a key element to building trust and confidence in decision-making.
The talent function can influence an organization’s most valuable assets—its people—and can help drive substantial value by engaging and retaining them. Building employees’ trust is important to activating long-term usage of AI-based tools in talent management, and the talent leader is expected to play a pivotal role in implementation. As one participant highlighted, the talent leader can either hinder adoption due to a lack of understanding or trust in the technology, or they can facilitate it by finding ways to enhance the work experience for both employees and managers. Organizations led by the latter will likely gain a competitive edge in the long term.
In August 2024, the Deloitte Center for Financial Services conducted a four-day online executive forum with 38 senior executives from some of the largest financial services institutions across the banking and capital markets (with annual revenues exceeding US$1 billion), commercial real estate (annual revenues exceeding US$100 million), insurance (annual revenues exceeding US$1 billion), and investment management (with annual revenues exceeding US$500 million) sectors. This group included talent executives, as well as a few from IT and strategy, all of whom had the authority to make decisions regarding the use of AI in talent management.