AI, Data, Talent Pressures Redefine Organizational Strategy

AI, talent pressures, and shifting work models are reshaping organizations as leaders rethink data, skills, and HR strategies across the Americas. Epsilon hosted global strategist and author Rishad Tobaccowala for a conference examining how organizations are restructuring work, talent strategies, and data use amid rapid technological and demographic shifts. 

The session brought together leaders navigating changes that are reshaping labor markets across Mexico, Latin America and the United States, particularly as AI becomes embedded in daily operations and workforce expectations continue to evolve.

During the event, Tobaccowala argued that the companies most likely to succeed will be those that understand the balance between human capability and digital systems. “Organizations that thrive will be those that combine human judgment with meaningful data, not those that rely on data alone,” he said. His remarks reflected concerns raised by the International Labour Organization (ILO), which has warned that AI in HR is expanding faster than the safeguards needed to manage its limitations. The ILO points to structural risks involving reductionist metrics, biased datasets, and opaque algorithmic programming, creating a complex environment for HR leaders.

These challenges come as studies across the Americas reveal persistent gaps between employer and employee perceptions of engagement, readiness and leadership effectiveness. TriNet’s 2025 State of the Workplace report highlights that 47% of employers believe their workforce is “extremely engaged,” while only one-third of employees agree. Mexican labor studies from Buk and Michael Page similarly show that organizations struggle to fill key roles due to the pace of technological change and the slow adaptation of education systems. Tobaccowala says that organizations need to redesign incentives, workflows, and internal mobility to keep pace with emerging demands.

During the event, he described how hybrid work, dispersed teams, and digital tools have changed decision cycles and team dynamics. Automation continues to accelerate, but its long-term value depends on how organizations integrate it into human-led processes. Tobaccowala told attendees that AI requires active oversight. “What you tend to do is have humans in the loop. You let AI do some of the tasks but then you put humans in the loop, including looking at what AI rejected to see where it made mistakes,” he said. He added that expecting AI to be neutral is unrealistic. “I do not think it is possible to have AI free of bias because I do not believe any human is free of bias,” Tobaccowala tells MBN. 

This observation aligns with the ILO’s view that AI often amplifies existing inequities. The organization warns that biased or incomplete data — often prevalent in markets with high informality, such as Mexico — can produce discriminatory outcomes, from skewed recruitment pipelines to compensation imbalances. Studies from OCC and regional platforms show that automated screening, targeted advertising, and predictive modeling can inadvertently mirror historical inequalities. Yet several HR technology providers in Mexico argue that supervised algorithms may help reduce human bias by expanding candidate pools and standardizing early-stage filters.

Tobaccowala also spoke about the rising role of HR within companies. “AI is going to make HR more important because they are going to have to retrain all their people,” he told MBN. He added that HR departments will need to reassess processes and benefits to align with shifting expectations, especially as organizations adopt AI-driven systems for screening, onboarding, and performance management.

His perspective resonates with broader labor trends across the Americas. TriNet’s research shows that AI adoption is widespread — 94% of employers and 84% of employees in the United States now use AI tools on the job. Yet employers remain hesitant to formally recognize AI skills in job descriptions or hiring criteria, despite employees reporting improved productivity. Industry analysts describe this as a “recognition lag,” reflecting uncertainty around how to define AI literacy in the workplace. In Mexico, similar gaps appear as companies increasingly use digital platforms for recruitment, internal communication, and benefits administration.

Beyond skills, expectations around flexibility and well-being remain central. Most employers surveyed in the TriNet report now consider three days in the office to be the optimal hybrid model, even though only 14% of employees agree. Regional data from Bonda shows that traditional benefit packages are losing relevance, with employee satisfaction falling from 79% in 2020 to 56% in 2024. Companies are shifting toward personalized benefits, mental health support, and child-care solutions to address these gaps.

Leadership effectiveness also emerged as a concern. Bonda found that only 27% of employees consider their leaders highly effective, highlighting the need for new leadership models centered on empathy, adaptability, and emotional intelligence. These findings align with Tobaccowala’s emphasis on the human elements of work, including identity and talent. When asked whether standardized recruitment criteria risk making people replaceable, he agreed that future hiring models will need to take individuality into account. “That is what we will focus on in the future,” he said.

As organizations confront technological and structural uncertainty, risk management is becoming central to business planning. Tobaccowala highlighted the importance of decision protocols, realistic timelines, and integration across functions. He noted that uncertainty is no longer an exception but a defining feature of contemporary markets. The ILO’s analysis reinforces this by emphasizing that transparency, data governance, and representation — not algorithms alone — determine whether AI reduces or reinforces inequity.

Tobaccowala called for balance between innovation and human judgment. “The companies that will lead are those that balance people and data, innovation and discipline, and long-term vision with short-term action,” he said. Epsilon’s conference provided leaders in Mexico and Latin America with a framework for aligning technology deployment with evolving workforce expectations, organizational design, and ethical considerations as AI continues to reshape how work is defined and performed.