AI at work: new data, new challenges for leaders.

The narrative around AI and jobs is evolving from one of straightforward automation to a more complex picture. Many companies are adopting AI primarily to boost productivity or revenue rather than to cut staff ([1]). In fact, a recent global survey found that AI investments had only a modest net negative effect on employment over the past year (about 5% more businesses reducing headcount than increasing it due to AI) ([2]) – a far cry from the mass unemployment some feared. And while certain support roles and routine tasks are indeed being automated, the immediate impact on overall job numbers has been muted.

At the same time, AI is creating entirely new jobs and demand for skills. Contrary to doom-and-gloom predictions, LinkedIn’s data shows that AI has added roughly 1.3 million new roles worldwide in areas like AI engineering, data analysis, and machine learning operations ([3]). Job postings for “AI-enabled” expertise are skyrocketing; for instance, listings seeking “AI agent” skills jumped by 1,587% in 2025 as companies sought talent to integrate AI into their business workflows ([4]). Roles that barely existed a few years ago – AI prompt engineers, machine learning specialists, AI ethicists – are fast becoming standard positions across industries ([5]).

The nature of work within existing jobs is also shifting. Rather than replacing professionals outright, AI is taking over specific tasks (like drafting reports or reviewing data), allowing employees to focus on higher-value activities. Still, early signs of disruption have appeared at the entry level. One rigorous study found that from 2022 to 2025, entry-level employment in highly AI-exposed fields (such as software coding and customer service) fell by roughly 13–20%, even while employment for more experienced workers in those fields grew during the same period ([6]). Business leaders worry that by using AI for junior-level tasks, they may undermine future talent pipelines. As one manager put it, companies are increasingly “utilizing AI where they would have recruited a young person… We’re going to skip that start point for people” ([7]). This highlights a new imperative: if AI handles more basic work, leaders must find other avenues to train and mentor early-career employees so critical skills aren’t lost.

Forward-thinking organizations recognize that success with AI isn’t about man versus machine – it’s about redesigning work so that people and AI can excel together. The conversation is shifting from “Will AI take our jobs?” to “How will our jobs change?” ([1]) as businesses explore new human-AI workflows. In many companies, these hybrid teams are quickly becoming the norm: according to Microsoft’s latest Work Trend Index, 49% of all Microsoft 365 Copilot AI interactions now involve complex cognitive tasks like analysis and writing, not just simple Q&A ([2]). And 58% of employees using AI say they’re already accomplishing work they could not have done a year ago without these tools ([3]). When skilled humans direct AI tools – rather than compete against them – the result can be higher productivity and even entirely new capabilities.

However, adopting AI tools is the easy part; the hard part is adapting how work is organized. Microsoft researchers call this the “Transformation Paradox” ([4]): organizations are eagerly rolling out AI technologies, but many have yet to fundamentally reconfigure processes, roles, and team structures to realize AI’s full value. Only 26% of AI users say their leadership is consistently aligned on a clear AI strategy ([5]). Similarly, a Deloitte study found that while executives overwhelmingly agree redesigning work for human-AI convergence is critical for ROI, only 6% of leaders feel they’re making significant progress on that front ([6]). The majority of firms remain stuck in pilot projects or siloed tech initiatives, with no blueprint for integrating AI into day-to-day workflows. That means lost opportunities – if you don’t rethink jobs and processes, as one expert warns, you’re “leaving most of the value on the table” ([7]).

On the flip side, organizations that do embrace comprehensive work redesign are pulling ahead. By empowering employees to work alongside AI, these leaders are starting to unlock outsized gains. One survey found that companies with AI fully integrated into operations are nearly four times more likely to report AI-driven revenue growth than those just experimenting at the pilot stage (58% versus 15%) ([8]). They are seeing tangible benefits – faster decision cycles, higher productivity, innovative services – because they paired technology adoption with changes in job design, team structure, and training. These examples underscore that AI transformation is as much about organizational design as it is about algorithms.

In some cases, AI is even prompting companies to reassess their management structures. Gartner analysts predict that by the end of 2026, 20% of large enterprises will use AI to flatten their hierarchies, potentially eliminating over half of their middle-management roles ([9]). Automation can handle tasks like routine reporting, scheduling, and basic performance monitoring – duties that traditionally occupied many supervisors. The efficiency gains could be significant, but so are the new leadership challenges: remaining managers must shift toward more strategic, value-added work, and companies will need to find new ways to develop future leaders if entry-level and middle-management ranks shrink. The takeaway for executives is that organizational agility and clear role design become even more critical in an AI-enabled workplace.

On the front lines of this AI transition are employees who are experiencing a mix of excitement and fear. A recent U.S. poll found that 50% of workers have already used generative AI tools in their jobs at least occasionally ([1]), evidence that AI is quickly becoming part of everyday work. Yet anxiety is running high: about one in five workers globally still believes their job won’t be affected by AI at all, while nearly half suspect AI will mostly benefit their employer at their expense ([2]). This gap between complacency and concern highlights the importance of clear communication and support from leadership. If executives fail to explain how AI will change roles – and how they will help their people adapt – they risk a backlash or, worse, a disengaged workforce.

The rapid pace of change is another source of strain. By one estimate, a third of employees went through 15 major changes to their work in the past year ([3]). Not surprisingly, many people are feeling change fatigue. Trust in senior leadership is wavering – dropping to around 72% of workers, down from 77% as recently as 2024 ([4]) – especially among younger staff. Instead, 60% of employees now look to their direct managers for support and stability amid uncertainty ([5]). This underscores how crucial middle managers are in guiding teams through AI-driven transitions, even as those very roles face new pressures.

Meanwhile, a serious skills gap is looming. External demand for AI-savvy talent is far outpacing internal skill development: as of this spring, postings for jobs requiring AI skills jumped 144% year-over-year ([6]), vastly exceeding overall job posting growth. Yet many companies have not invested adequately in upskilling their people to meet this demand. In fact, 34% of finance leaders admit that employee training is the most underfunded part of their AI budget ([7]). As one report warns, too many firms are “treating AI like an IT project” – pouring money into technology without enough support for the humans who must use it – leaving the workforce “watching AI arrive without knowing what to do with it” ([8]). The result is a double hit: underprepared employees become frustrated or fearful, and the organization fails to capture the full value of its AI investments.

The good news is that employees are eager to adapt – and companies can build on that enthusiasm. In a global survey, 65% of workers recognized the need to acquire new AI-related skills, and 52% are already taking steps to upskill themselves to future-proof their careers ([9]). Forward-looking organizations are starting to support this effort by establishing internal AI academies and providing clear guidelines for using AI responsibly ([10]). By treating AI adoption as an ongoing learning journey – not just a one-off software installation – leaders can empower their people to thrive alongside new technologies. This kind of investment in human capability doesn’t just alleviate fear; it ensures that the organization has the skills and buy-in needed to fully leverage AI’s potential.

As AI initiatives accelerate, many companies find their leadership and governance approaches lagging. Roughly three-quarters of corporate boards have now greenlit major AI projects, yet nearly half have not set clear policies for AI oversight or integrated AI risks into board-level monitoring ([1]). It’s not surprising, then, that 78% of business executives lack confidence their organization could swiftly pass an independent audit of its AI systems’ accountability ([2]). In other words, many organizations are rushing into AI without the guardrails and clarity around responsibility needed to manage it safely.

This isn’t just a compliance issue – it’s a cultural one. In one study, 60% of executives said they already use AI for important decisions, but only 5% believe their company handles AI-related risks effectively ([3]). More than half (56%) admit their AI initiatives focus solely on efficiency and business outcomes, whereas only 40% prioritize both business and human impacts ([4]). Small wonder that 65% of organizations say their culture needs to change significantly to adapt to AI’s influence on work ([5]). To lead an AI-powered transformation successfully, executives must double down on ethics, transparency, and employee engagement – or risk eroding trust and facing internal pushback.

Increasingly, external stakeholders are stepping in to help set those guardrails. Labor unions, for example, have begun negotiating AI provisions to protect workers’ interests. The NewsGuild, a major media union, now has AI clauses in roughly 85 to 90 of its contracts ([6]) – obliging employers to notify employees when introducing AI, define the boundaries of AI’s use, and allow workers to contest unjust outcomes ([7]). Even in Hollywood, the SAG-AFTRA actors’ union made history in 2026 by securing strict limits on the use of “digital doubles” and other AI tools that replicate performers, requiring consent and compensation for any AI-generated replicas of an actor’s likeness ([8]). These agreements, driven by employee concerns, provide a blueprint for introducing AI ethically – through transparency, negotiation, and respect for human impact.

Regulators, too, are catching up with AI’s advance. In the European Union, a sweeping AI Act has classified many workplace AI applications – from hiring algorithms to productivity monitoring – as “high risk” systems that demand rigorous oversight ([9]). While the EU recently extended certain compliance deadlines from August 2026 to late 2027 for existing AI systems ([10]), key transparency measures still took effect this month. Companies must now inform employees when AI is being used in processes like recruiting or performance evaluation ([11]). Similar legal and policy moves around the world – from proposed laws on AI in hiring to new AI governance guidelines – signal that organizations will be expected (and often required) to integrate ethics, accountability, and worker protections into their AI strategies.