AI’s Workforce Shake-Up: Hard Truths & Surprises

New research offers a stark snapshot of how AI is both eliminating and creating jobs. A global survey of recruiters released this week found that one in three companies is already replacing entry-level roles with AI technologies ([1]). In industries like technology and manufacturing, the figure is even higher—around 40% of employers report using AI to automate junior positions instead of hiring new graduates ([2]). This data confirms that the lowest rungs of the career ladder are being chipped away by automation, fueling concerns for Gen Z and others trying to enter the workforce.

Yet, counterintuitively, AI isn’t causing mass layoffs everywhere. A recent U.S. Chamber of Commerce survey of small businesses revealed that 82% of those adopting AI actually increased their staff over the past year ([3]). Rather than cutting employees, these smaller firms used AI to automate tedious “overhead” tasks while redeploying human workers to higher-value duties like customer service and creative problem-solving ([4]). In other words, AI became a tool for growth and productivity, enabling people to focus on work that drives revenue and innovation instead of drudge work.

Real-world examples from the past few days highlight how the impact of AI on jobs is more complex than simple automation. Several large companies that tried aggressive AI-driven layoffs are now backtracking. Automaker Ford, for instance, is rehiring hundreds of experienced engineers to fix quality problems that arose when certain tasks were automated—problems the AI couldn’t adequately address ([5]). Similarly, after replacing 40 customer service representatives with an AI voice bot, Australia’s largest bank found the bot struggled with customer needs; call volumes spiked and the bank reversed those job cuts ([6]). That reversal was even hailed as a 'massive win' by the bank’s employee union, reflecting the human cost of mismanaging an AI transition ([7]). These turnabouts make one thing clear: the net effect of AI on employment is not straightforward. The technology can boost efficiency, but overzealous cuts can quickly boomerang if AI falls short of human capabilities.

Amid the disruption, a growing number of forward-looking organizations are choosing to adapt rather than trim their workforce. They recognize that realizing AI’s full value often requires rethinking jobs and processes, not just installing new software. Tech giant Microsoft, for example, is transforming its operations by redesigning workflows and focusing on upskilling employees, viewing AI adoption as a holistic business transformation instead of a typical IT rollout ([1]). This means methodically weaving AI into every team’s daily work while training staff at all levels to leverage these tools.

New industry research backs this people-centric approach. A recent analysis by Boston Consulting Group found that “future-built” companies—those unlocking AI’s potential most effectively—plan to upskill more than 50% of their employees on AI, versus only 20% in lagging firms ([2]). These leading organizations are also four times more likely to establish structured AI learning programs and set aside time for employees to train on new AI tools ([3]). In practice, this investment in learning pays off: employees at such companies tend to view AI as an empowering assistant in their work rather than a threat, and the organizations see faster scaling of AI initiatives and greater business value generated.

Salesforce offers a vivid case study of how to integrate AI while prioritizing people. In 2025 the company introduced an AI agent (“Agentforce”) to handle customer support queries, and within months the bot was resolving 2.6 million conversations with a 63% success rate—matching human agents on customer satisfaction ([4]). Instead of using this as a reason to cut headcount, Salesforce reimagined the roles of its support staff. Over the past year, it retrained and redeployed hundreds of support engineers into new, fast-growing roles across the company, while choosing not to backfill their old positions ([5]). Those workers carried their deep product and client knowledge into departments like sales and product development, where their contributions led managers to call them some of the best hires they’d ever made ([6]). Salesforce has formalized this process into a “4R” framework—Redesign work processes, Reskill every employee in using AI, Redeploy talent flexibly to higher-impact areas, and Rebalance workloads between humans and AI. As one executive put it, the key is investing in people’s ‘next chapter’ as much as in the technology itself ([7]). By betting on existing talent and helping them grow, companies can turn automation into an opportunity for employees to take on more fulfilling, valuable work.

The human side of AI-driven change is proving to be challenging. Many employees are fearful about what rapid AI adoption means for their careers and well-being. According to a global survey, only 38% of workers believe their organization can effectively adapt to technological disruptions, and just 30% feel confident about how their leaders will handle the workforce changes brought by AI ([1]). Fewer than half of employees now feel secure in their jobs—a sharp drop from 59% who felt secure just a year ago ([2]). This erosion of trust and psychological safety poses a real threat to morale and retention if not addressed.

In the absence of clear guidance from employers, some employees are taking a cautious, under-the-radar approach to AI. A Microsoft study revealed that 78% of employees using AI at work are “bringing their own AI” tools (often without IT’s knowledge) to help with their jobs ([3]). And over half of these workers worry that using AI for important tasks could make them look replaceable to their managers ([4]). In other words, people are turning to AI to boost productivity, but many feel they must hide it, fearing it could paint a target on their backs. This quiet, unguided adoption of AI can create security and ethical risks—and signals that employees crave more clarity and reassurance from leadership.

One way to rebuild trust is to double down on the uniquely human skills and roles that AI cannot replicate. In practice, employers are starting to place greater emphasis on attributes like critical thinking, creativity, and empathy—the skills that allow workers to interpret AI outputs, spot errors, and connect with customers on a human level. In fact, 73% of talent acquisition leaders now rate critical thinking as the number-one skill they seek in new hires, specifically because those hires will need to evaluate and correct AI-driven results when necessary ([5]). Companies are also creating new job roles that leverage human strengths alongside AI: for example, at Salesforce some frontline support agents whose routine tasks were automated have been retrained for a new “forward-deployed engineer” position focused on higher-level problem solving in tandem with AI tools ([6]). Even managers are being retrained to lead hybrid human–AI teams, with one leader noting that while AI "provides objectivity," managers must still "bring context"—and that the 'human element remains critical' for success ([7]). By emphasizing what humans do best and involving employees in co-creating new roles, organizations can alleviate fear and build a more resilient, innovative workforce.

All of these factors are putting AI on the agenda in every boardroom and government capital. In one recent study, 98% of executives globally said they plan to change their organization’s design within two years to better leverage AI, and 99% expect to see at least some reduction in headcount as a result ([1]). Boards of directors, eager to see returns on costly AI investments, are pressing leadership teams to aggressively reorganize workflows, flatten hierarchies, and rethink job designs around human–machine collaboration ([2]). The mandate from the top is clear: scale up AI-driven efficiency, but do so in a way that delivers real business value.

However, a gap is emerging between leadership ambitions and on-the-ground readiness. Only 11% of recruiting and HR leaders believe their executives are well prepared to guide their people through AI-driven change ([3]). Even many CEOs confess to uncertainty about execution: 79% of leaders in a Microsoft survey agreed that adopting AI is essential to competitiveness, yet 59% said they struggle to quantify AI’s return on investment and 60% worry their organization lacks a clear AI implementation plan ([4]). This misalignment can lead to hesitant decision-making at the very moment employees need decisive direction. Closing this gap will require leaders to articulate a compelling vision for how AI will improve the organization and to invest in managerial capabilities for leading through technology-driven shifts.

Finally, policymakers and labor stakeholders are stepping up involvement, which in turn pressures companies to be more proactive and transparent. On July 6, Illinois enacted a landmark AI accountability law—the first in the U.S. to mandate annual third-party audits of high-risk AI systems for fairness and safety ([5]). Notably, this state law had the support of major AI companies like OpenAI and Anthropic, which helped push it through the legislature ([6]). The fact that industry players are endorsing regulation signals a recognition that consistent standards may ultimately benefit everyone by building public trust ([7]). Meanwhile, worker organizations are also engaging: unions worldwide have made clear they expect a say in how AI is implemented, bargaining for guardrails around issues like surveillance and job security. When the Commonwealth Bank of Australia reversed its AI-driven layoffs, the finance sector union celebrated it as a 'massive win' for employees ([8]). In this environment, executives need to anticipate greater scrutiny of their AI plans—from regulators, investors, and their own people—and lead with transparency, ethics, and inclusivity. Those who navigate the transition with a human-centric approach won’t just avoid backlash; they’ll cultivate a more adaptable, future-ready organization.