AI at work: surprising new realities for leaders.

In the past year, a wave of layoffs tied to AI and automation has swept through multiple industries. Over 245,000 employees were laid off across the tech sector during 2025, and about 28.5% of those job cuts were directly attributed to companies implementing new AI systems . Early 2026 saw further reductions as some firms eagerly bet that generative AI tools could fully replace humans in roles ranging from customer support to software development.

([1])This week, however, brought fresh evidence that many of those AI-driven layoffs may have been premature. Multiple new reports reveal a striking "AI boomerang" effect: between roughly one-third and as many as two-thirds of companies that replaced employees with AI have since rehired for those same positions or expressed regret over the decision ([2]) ([3]). In plain terms, the rush to automate has often backfired, as businesses learned the hard way that technology didn’t always meet expectations.

([4])Real-world examples are driving this lesson home. Automaker Ford, for instance, is reportedly rehiring hundreds of engineers it laid off after finding that automated quality-control systems failed to catch certain design flaws that experienced humans would catch ([5]). Similarly, Commonwealth Bank of Australia reversed an AI-driven staff cut when its new call-center chatbot became overwhelmed by customer needs – a retreat the bank’s employee union celebrated as a 'massive win' for workers ([6]).

Even tech leaders have discovered the limits of what AI can do alone. IBM managed to automate about 94% of routine HR inquiries with AI chatbots, but the remaining 6% – often involving complex, sensitive issues – still required human judgment . The company ultimately restored those roles and went further, announcing plans to triple its intake of junior employees in 2026 after its chief HR officer warned that cutting entry-level hiring for AI would leave 'no pipeline' of future talent . These turnarounds all point to a common truth: human insight and creativity remain critical. In one survey, 40% of companies found AI couldn’t replicate employees’ institutional knowledge, 38% realized they needed more human oversight than anticipated, and 35% saw productivity gains fall short of expectations ([7]).

The focus is now shifting from what AI can do to how organizations must change to take advantage of it. A global Mercer study found that 98% of senior executives plan to restructure parts of their organizations within two years to better integrate AI, and 99% expect these technologies to result in at least some workforce reductions along the way . In short, boards and CEOs have recognized that simply layering AI tools onto existing processes won’t magically boost productivity – instead, they are coming to see AI adoption as a management and organizational challenge that demands rethinking workflows, roles, and decision-making norms .

However, there remains a wide gap between leadership ambition and on-the-ground reality. Only 32% of executives feel their workforce is currently effective at collaborating with AI systems , and a mere 6% say they have made significant progress in redesigning how humans and AI work together day-to-day . Meanwhile, Microsoft’s latest Work Trend Index highlights a "Transformation Paradox": employees are rapidly embracing AI in their daily work – 49% of Microsoft 365 Copilot users’ requests are now for writing, analysis and other complex cognitive tasks that used to require human expertise – yet just 26% of these AI-equipped workers agree that their leadership has a clear, unified vision for AI’s role, indicating technology is far outpacing organizational alignment . Without adapting management practices, culture, and training to support human–AI collaboration, much of AI’s promised value could remain untapped.

([1])Some leading organizations are responding with bold structural changes. Social media giant Meta, for example, has reshaped a major product group into small "AI-native pods," even rebranding employees as 'AI builders' and 'AI leads' to drive a flatter, faster AI-powered workflow ([2]) ([3]). Other companies are similarly streamlining reporting lines, establishing new AI oversight roles, and even treating certain algorithms as "virtual employees" that work alongside their human teams . As one KPMG report put it, the winners of the AI era will not be the companies that simply adopt AI first, but those that reorganize their operations and talent most effectively around AI’s capabilities .

All these developments put the spotlight on the human side of AI transformation. Employers understand that new technology demands new skills – they anticipate 39% of workers’ core skills will need to change by 2030 to keep pace with intelligent automation . And while AI will undoubtedly disrupt many roles, the World Economic Forum projects it will also create 170 million new jobs globally by 2030 even as 92 million are displaced – a net gain of 78 million positions that suggests work will be radically reshaped, not just eliminated .

The talent race is already on. The demand for AI skills has surged: in the U.S., job postings requiring AI expertise have jumped 144% over the last year . Workers who acquire expertise in areas like machine learning or data analytics are now commanding salary premiums up to 62% higher than peers (with some specialized tech roles seeing pay more than double) . And job titles that were virtually unknown just a few years ago – from prompt engineers to AI ethicists – are fast becoming commonplace as companies scramble to fill new capability gaps and harness AI responsibly .

Yet even as opportunities grow, workers are feeling the strains of this rapid change. Across sectors and regions, employees report rising anxiety about job security and the relentless pace of automation – in one recent survey, only 21% of workers said they feel confident using AI tools in their current role . This uncertainty has already taken a toll on morale: global 'thriving at work' well-being scores have plunged from 66% in 2024 to just 44% in 2026 amid the AI upheaval . For executives leading AI initiatives, this erosion of trust and engagement is not just a cultural issue but an emerging operational challenge – demoralized teams are less willing to embrace new AI-driven workflows, slowing the pace of change .

【10†L15-L2 ([1])rprise that labor groups and policymakers are stepping in to promote a more balanced, human-centric AI transition. The landmark 2023 Hollywood writers’ and actors’ strikes, which secured new limits on studios’ use of AI, have inspired unions in industries from media to finance to fight for contract clauses that protect jobs, mandate transparency, and require consultation before AI adoption. Governments are also moving swiftly on this issue. For example, as of August 2026, the European Union’s AI Act is being enforced with strict transparency rules – requiring AI systems to clearly disclose AI-generated content and when users are interacting with a machine rather than a person ([2]). These external pressures reinforce the need for business leaders to not only invest in AI technology, but also uphold ethical standards, adapt policies, and engage openly with their workforce – ensuring AI is implemented in a way that earns employee trust and meets regulatory expectations.