How AI is taking over jobs that can be split apart

Jobs with bundled tasks and their vulnerability to AI automation

AI is changing how we think about work. Not every job faces the same level of disruption; what matters is whether the tasks within that job can be separated. According to economists Luis Garicano, Jin Li, and Yanhui Wu, the more a job’s activities can be “unbundled”—meaning individual tasks can be split and handled independently, the greater the risk that AI will take over those functions. When tasks depend heavily on each other, involve continuous decision-making, or require accountability for outcomes, AI becomes less effective as a replacement and more valuable as a support tool.

For executives, this separation of tasks is more than an academic discussion. It affects how you design your organization, how you plan workforce transitions, and how you capture value from AI. In jobs where tasks are interconnected, think leaders, complex project teams, or roles with heavy responsibility, the human element is difficult to remove. These roles require context, emotional intelligence, and judgment that AI cannot yet replicate. In such cases, AI should be viewed as an amplifier of human performance, not a substitute.

The research framework from Garicano, Li, and Wu shows that the “cost of breaking the bundle” determines how far AI can go. When the cost of separating human and machine tasks is low, automation speeds up. When it’s high, because those tasks rely on shared context or real-time feedback loops, AI’s role remains supportive. For C-suite leaders, this signals that the challenge ahead is not about resisting automation but understanding where it naturally fits into your business architecture. The organizations that thrive will be those that identify the right balance between leveraging AI efficiency and protecting the human coordination that drives innovation and accountability.

High-risk creative and technical professions facing disruption

AI is advancing fastest in areas with repetitive, clearly structured tasks. Studies from Digital Planet at Tufts University show that roles such as writers and authors, computer programmers, and digital interface designers are among the most exposed. The numbers are substantial, 57% of writing and authoring roles, 55% of programming positions, and 55% of design roles could be impacted. These jobs often involve processes that can be defined and executed by pattern recognition systems and large language models, which explains the rapid uptake of automation in such fields.

For business leaders, this is a signal to re-evaluate workforce strategy. AI does not eliminate the need for talent, but it reshapes what that talent should focus on. Technical and creative teams will need to move toward higher-value work, designing systems, driving product vision, or integrating human insight with AI-generated output. Some professions, such as software developers, management analysts, and market research analysts, may experience income pressure, even if their roles evolve rather than disappear. The shift will be less about replacement and more about transformation.

Executives should consider how these changes affect both short- and long-term planning. Workforce flexibility, skill reinvention, and the ability to integrate human expertise with AI tools will define organizational resilience. Companies that treat automation as a partnership, allocating repetitive work to machines while empowering people to handle complex reasoning, will find themselves ahead. The key task now is identifying which parts of your value chain can be safely automated without weakening creativity, accountability, or quality.

AI does not level all industries equally. It targets task structures, not job titles. The leaders who build around that understanding will be the ones shaping the next productive era of work rather than reacting to it.

Key takeaways for decision-makers

Identify task structures to manage automation risk: Leaders should evaluate which roles in their organization consist of separable tasks, as these are most vulnerable to AI automation. Prioritize redesigning such roles to increase interdependence between human judgment and automated processes to preserve value and agility.
Reskill creative and technical teams for AI collaboration: Jobs in writing, programming, and design face significant AI-driven restructuring. Executives should focus on upskilling teams to work alongside AI tools, emphasizing creativity, strategy, and complex reasoning, areas where human expertise delivers competitive advantage.