AI’s workplace Shake-Up: efficiency boosts, talent gaps & the human factor.

For some companies, the push to integrate AI isn’t only about cutting costs – it’s about fundamentally changing who does the work and how. This week General Motors took the unusual step of laying off 600 members of its IT department (over 10% of that workforce) in what it termed a 'skills swap' ([1]). Rather than simply eliminate those positions, GM is recruiting new employees with 'AI-focused' skill sets to refill many of the roles ([2]). The company is seeking talent in areas like AI model development, data analytics, cloud engineering, and even prompt engineering – people who, as one insider put it, know how to build with AI from the ground up ([3]). In effect, GM is trading out staff whose expertise 'no longer fits' for workers prepared to construct the next generation of AI-driven systems.

Tech firms that have invested heavily in AI are likewise reshaping their organizational structures. Meta’s latest restructuring not only cuts headcount; it’s also inventing new positions such as 'AI builder', 'AI pod lead', and 'AI org lead' ([4]). Teams are being reorganized into nimble, AI-centric "pods," often by migrating employees into a central AI-focused division, as the company aims to 'fundamentally rewire how we operate' for greater productivity ([5]).

Other businesses are even experimenting with radical team models built around AI. At the freelancing platform Upwork, the CEO decided to cut 25% of the workforce in a bid to move faster, arguing that 'AI means smaller, differently resourced teams in product and engineering can make a bigger impact than ever' ([6]). In her message to employees, she outlined plans to consolidate redundant tasks, collapse layers of management, and build flatter, more agile groups with AI boosting their capacity ([7]). Meanwhile, crypto company Coinbase has gone so far as to eliminate most traditional managers altogether. In a bold May reorg accompanying a 14% staff reduction, CEO Brian Armstrong introduced 'AI-native' pods that include single-person teams using AI agents to perform what used to be done by entire departments of engineers, designers, and product managers ([8]).

As AI systems begin handling more knowledge work, companies across industries are re-imagining processes to capitalize on these capabilities. The World Economic Forum’s recent report highlighted case studies where AI tools drastically reduced workloads – in one, an AI system analyzing tax and regulatory data uncovered $120 million in savings and cut a filing process from weeks to days, while another firm’s lab automation bot shrank a 30-minute ordering task to mere seconds, freeing up 30,000 hours of work per year ([9]). Leaders also note that automating rote tasks can improve employee well-being: early adopters report AI has reduced repetition and burnout, allowing staff to focus on higher-value work and creativity ([10]). And looking to the near future, some foresee AI becoming an official 'team member' itself – ServiceNow’s Chief Strategy Officer Hala Zeine has predicted that org charts will soon include AI agents as formal team members alongside humans, each with defined responsibilities and performance metrics ([11]).

For many workers, the impact of AI on their day-to-day jobs has been limited so far – but that hasn’t eased anxieties about what’s coming. In a recent survey of over 4,000 U.S. professionals, 75% said they have not experienced any changes to their job due to AI yet ([1]). Only 6% reported that their own role has changed because of AI. Yet nearly half (42%) of these workers fear that AI could eventually threaten their job security ([2]). In short, there’s a wide gap between the constant talk of AI disruption and most employees’ current reality.

At the same time, a growing number of employees are proactively experimenting with AI – sometimes under management’s radar. Research shows that while only about 40% of organizations have officially invested in advanced AI tools like large language models, workers at over 90% of those companies report using their *own* AI tools on the job ([3]). This shadow adoption can be a double-edged sword. Roughly half of U.S. employees use AI at work without knowing whether it’s allowed, and 44% knowingly use it in ways their employers haven’t sanctioned ([4]). On one hand, these self-starters are boosting their individual productivity with AI and sharpening new skills – for example, more than a quarter of workers say AI motivates them to learn new capabilities faster ([5]). On the other hand, unauthorized AI use creates risks around errors, data security, and compliance if organizations don’t catch up.

Where AI is arriving openly in the workplace, it’s also provoking resistance and calls for safeguards. Across industries, labor unions are emphasizing demands for transparency, fairness, and human oversight in AI deployment ([6]). In the media sector, the union at ProPublica went so far as to stage a 24-hour strike in April, pushing for contractual guarantees on job security in an 'AI-driven industry' and a say in if and how new AI tools are used in reporting ([7]). Journalists at outlets like The New York Times and the Associated Press have likewise protested what they call "woefully inadequate" AI policies after incidents such as an AI-generated book review that mimicked another publication’s writing ([8]). Meanwhile, in the tech world, nearly all of Google DeepMind’s UK staff – 98% – voted this month to unionize, motivated in part by concern over their employer’s involvement in military AI projects and a desire for stronger ethical governance of AI work ([9]).

These reactions are a warning for leaders: failing to address employees’ concerns can derail an AI initiative. A labor relations analysis found that employers who treat AI purely as a technical implementation – rather than also addressing the human and trust aspects – are far more likely to encounter workforce resistance ([10]). Conversely, organizations that communicate early, involve employees in the rollout of AI changes, and maintain clear human oversight tend to see far less pushback ([11]). In this delicate moment, a people-first approach isn’t just about avoiding blowback; it’s key to unlocking AI’s potential productivity gains in a sustainable way.

The race to embrace AI is exposing critical capability gaps – both in the workforce and in the boardroom. Even as automation displaces certain jobs, companies are struggling to fill new tech-heavy positions. In the U.S., roughly 275,000 AI-related job postings were open at the start of this quarter, even as the industry was experiencing record layoffs ([1]). Employers report a 92% surge in hiring for AI roles in 2026, often with salaries carrying a 50%+ premium for in-demand skills ([2]). Yet the workers being laid off – in areas like customer support, routine administration, and middle management – typically don’t have the machine learning and data science skills needed for those open roles ([3]). This stark skills mismatch means that without significant retraining, many displaced employees simply can’t step into the new jobs AI is creating.

Organizations are adopting different strategies to close this gap. Many are in a talent 'arms race,' either hiring externally or trying to rapidly upskill their people. Automaker GM’s 'skills swap' is one example of bringing in fresh AI expertise following layoffs ([4]) ([5]). Meanwhile, some tech giants are willing to pay unprecedented sums to attract top AI experts – one report noted that Meta offered compensation packages up to $1.5 billion for a single elite machine-learning engineer ([6]). Other companies are focusing more on developing their existing workforce. For instance, at the World Economic Forum’s 2026 meeting in Davos, 25 major firms – including Cisco, Wipro and ServiceNow – collectively pledged to train over 120 million people worldwide in digital and AI skills by 2030, helping create new 'AI‑native' roles for the future ([7]). These kinds of large-scale upskilling initiatives reflect a commitment to avoid leaving existing employees behind.

Lastly, there’s the challenge of leadership itself adapting to the age of AI. A recent KPMG report revealed nearly three-quarters of corporate boards have only 'moderate or limited' understanding of AI technologies and their implications ([8]). To address this, KPMG and INSEAD have introduced new AI governance principles to guide board directors in providing effective oversight as AI transforms their organizations’ strategies and risk profiles ([9]). Even among executives, direction can be inconsistent: 75% concede their company’s AI strategy is 'more for show' than actual substance ([10]). And 79% of enterprises say they face significant challenges turning AI investments into real business value – with 54% of C-suite leaders admitting that the process of adopting AI is actually 'tearing their company apart' internally ([11]). As the World Economic Forum points out, AI transformations more often stumble due to people and process issues rather than technology failures ([12]). The organizations that succeed are typically those that deliberately redesign workflows and roles to mesh human expertise with AI capabilities, invest in training and change management, and foster a culture of trust and adaptability. The bottom line for leadership: effectively leveraging AI is not just about buying the latest algorithms – it’s about preparing your people, and yourselves, for a new way of working.