The Pandemic Was a Dry Run

The COVID-19 pandemic permanently altered work organization by normalizing hybrid and remote working, and this shift is now being deepened by the rapid adoption of artificial intelligence (AI). Firms increasingly deploy AI to automate routine tasks and reorganize workflows, with hybrid work serving as an adjustment mechanism that can mitigate job loss by redistributing work rather than eliminating roles. Evidence shows that hybrid arrangements have remained remarkably stable despite return-to-office (RTO) pressures, while AI use among workers is rising faster than firms’ formal implementation strategies. The productivity effects of remote work are uneven—benefiting startups through expanded access to global talent but posing challenges for more rigid incumbent firms—and access to flexibility is unequally distributed across demographic groups and sectors.

˜˜˜˜˜˜

The COVID-19 pandemic functioned as a large-scale natural experiment in remote work. What initially emerged as an emergency response rapidly evolved into hybrid work arrangements—typically three days in the office and two days at home. While working from home (WFH) was initially mandated, it has since become a strategic choice for firms and workers. This choice is now being reshaped by the diffusion of artificial intelligence, which is transforming task allocation, organizational design and labor demand.

As AI systems increasingly assume routine and administrative tasks, several corporations—particularly those with largely remote workforces—have restructured workflows accordingly. In this environment, hybrid work may act as an adjustment margin that mitigates outright job loss by reallocating work time rather than eliminating positions. Workers increasingly collaborate with AI tools that augment productivity, allowing them to redirect effort toward higher-value tasks, retraining or nonmarket activities such as community engagement, creative pursuits and volunteer work.

This transition raises an important question: Are workers choosing remote or hybrid work, or are they responding to organizational mandates driven by AI adoption?

This transition raises an important question: Are workers choosing remote or hybrid work, or are they responding to organizational mandates driven by AI adoption? To the extent that AI displaces tasks previously performed on-site and reduces labor’s share in production, hybrid work arrangements may help smooth labor-market adjustment by spreading work over fewer days to avoid outright job loss.

The concept of “digital coworkers” illustrates this shift. AI assistants—or “digital twins”—are increasingly deployed to manage email, consolidate messages and even attend meetings. Zoom’s chief executive officer (CEO), Eric Yuan, suggested that such digital twins will work alongside employees, handling routine coordination tasks (Abril, 2026). Rather than replacing workers, these tools will reshape the division of labor within firms.

Survey evidence underscores the rapid diffusion of AI in the workplace. By August 2025, 23 percent of US workers reported using AI tools several times per week, up from 11 percent in late 2024. Consolidating information and generating ideas was the motivating factor for more than 40 percent of workers; automation of basic tasks, such as AI writing and editing tools, represented 30 percent; and collaboration was important for only 11 percent of workers. In a Gallup poll, only 22 percent of respondents felt that their workplace organizations had clearly communicated a plan for integrating AI. Employees were less likely to view AI with suspicion if their managers clearly articulated the implementation process, outlining the benefits and possible risks (Gallup Workplace, December 2025).

Despite high-profile return-to-office mandates, hybrid work has proven remarkably persistent.

Despite high-profile return-to-office mandates, hybrid work has proven remarkably persistent. Data from the American Time Use Survey (ATUS) indicate that approximately one-third of workers reported performing at least some work from home, while 25 percent of all paid workdays between 2023 and 2025 were fully remote (Buckman et al., 2025). Gallup surveys have shown that 50–56 percent of full-time US employees hold remote-capable jobs. These hybrid workers spend approximately 45 percent of their time on-site—equivalent to two to three days per week—and this share has remained stable since the pandemic (Gallup Workplace, September 2025).

The productivity implications of remote work differ markedly across firms. Using an instrumental variables (IV) design based on pre-pandemic occupational remote-work suitability, Gupta, Gupta and Simintzi (2025) found that remote work increased the productivity of startups while reducing it for established firms. Startups benefit from access to a broader global talent pool, alleviating hiring constraints that traditionally disadvantage young firms. Established firms, by contrast, often operate within rigid organizational structures and legacy workflows that are less adaptable to remote arrangements. Gupta et al. wrote: “The effects on productivity are larger for startups which are fully remote, instead of hybrid, which are therefore able to access an even broader geographic market for talent. We find that remote startups are able to hire workers for lower salaries, suggesting greater labor access, and that the productivity effects are driven by geographic markets with weaker baseline access to talent.”

Hybrid work is unevenly distributed across demographic groups and industries. Younger workers enjoy greater autonomy: Approximately 66 percent of Generation Z (born after 1996) workers reported having a choice over which days they work on-site, compared with just 27 percent of Millennials (1980-1996) (Gallup Workplace, July 2025). Sectoral differences were even more pronounced. Only 9 percent of technology workers regularly attended the office, compared with 46 percent of federal government employees (Gallup Workplace, September 2025).

Views among business leaders regarding work arrangements have evolved alongside AI adoption. While executives such as David Solomon of Goldman Sachs and Elon Musk of Tesla previously advocated five-day office attendance, JPMorgan Chase CEO Jamie Dimon has argued that AI deployment could reduce the workweek to three or four days over the next two decades (Abril, 2026). Even firms with strict on-site policies may face pressure to adjust as AI implementation requires retraining, task reallocation and, in some cases, early retirement.

Coordination, monitoring and the role of offices

The notion of a single, centralized workplace has also eroded. Teams are increasingly distributed across multiple locations. The share of remote-capable employees reporting that their team is spread across different worksites rose from 13 percent in 2023 to 27 percent in 2025 (Gallup Workplace, September 2025). In this context, collaboration relies less on informal, colocated interactions and more on structured digital communication, such as email.

The principal method of communication in the hybrid-work environment is email. But is email a substitute for in-person communication? Paradoxically, email complements on-site work in communication between a boss and a worker, while it substitutes for in-person communication between workers.

Return-to-office mandates have been predicated on the inability to coordinate and the inaccurate monitoring of work. Using data from email communication, Prithwiraj (Raj) Choudhury et al. (2026) found that email in vertical communication (across pay scales) and on-site work were complements due to increased monitoring efficiency. Bosses can meet with employees to discuss performance while formalizing the discussion in writing. Email messages substitute for horizontal communication between workers at the same pay grade by specifying tacit knowledge, so workers do not need to collocate. Information gleaned from socializing outside of the workplace is written in digital communication.

Spillovers to local economies

One of the most visible spillovers of hybrid and remote work has been its impact on urban service sectors, particularly restaurants. Middle-market restaurants face mounting pressure from rising labor costs, increased law-enforcement presence and reduced weekday foot traffic. In the Washington, D.C. metropolitan area, 92 restaurants closed in 2025, compared with 73 closures the previous year (Carman, Heil and Rojas, 2025). While remote work is not the sole driver of these closures, it has amplified existing vulnerabilities in sectors dependent on dense, office-centered demand.

AI and labor reallocation

Worker displacement remains the most visible manifestation of AI adoption. In late 2025, Amazon announced the elimination of 14,000 corporate jobs, Google’s YouTube proposed voluntary resignations, and Salesforce cut approximately 4,000 customer-support roles (Abril, 2025). However, some of this displacement reflects task reconfiguration rather than net job loss, as organizations redesign roles to meet evolving skill demands (Walsh, Gartner Group, as cited in Abril, 2025).

Key takeaways

• The pandemic accelerated a structural shift in work—hybrid work is highly persistent despite return-to-office pressure.
• Hybrid-work access is higher amongst Zoomers than Millennials and represents a new form of labor-market privilege.
• Offices are no longer coordination hubs; they are monitoring hubs.
• Local service economies are bearing the real costs of hybrid work.
• Artificial intelligence is reshaping work more through task reallocation than outright job loss.

Conclusion

The pandemic was not merely a temporary disruption to work arrangements; it was also a rehearsal for a deeper transformation driven by hybrid work and AI adoption. Together, these forces are reshaping labor markets, firm dynamics and urban economies. Policies that support worker retraining, clarify AI implementation strategies and facilitate adjustments in local service sectors will be critical to ensuring that productivity gains translate into broadly shared economic benefits.

ABOUT THE AUTHOR

Dr. Swati Bhatt has been on Princeton University’s Economics faculty since 1992. Her research interests include the economics of digitisation and industrial organisation with a focus on the technology industry. Previously, she worked for the Federal Reserve Bank of New York (FRBNY) and taught at the New York University (NYU) Stern School of Business.