Business

Planning for Generative AI Means Tracking Tasks, Not Predicting Layoffs

ILO research suggests generative AI will more often transform and augment work than eliminate whole occupations, making task-level planning essential for newsrooms.

Editorial note: This original explainer was prepared with AI assistance and human editorial review from the source documents cited below. It is not personalized professional advice and does not report a new event.

Generative artificial intelligence is often discussed through dramatic questions: Which jobs will disappear? How many workers will be replaced? For newsroom leaders, those questions are too blunt to guide responsible planning. Research from the International Labour Organization (ILO) points toward a more useful approach: examine how particular tasks may change, then plan for redesigned work, consultation, and safeguards.

The distinction matters because the ILO’s measures of “exposure” and “potential” describe the extent to which tasks or occupations may be affected. They are not forecasts of layoffs, job losses, productivity gains, or outcomes for any specific newsroom. A responsible plan therefore starts with evidence and local review rather than turning a global estimate into a headcount decision.

What the ILO is measuring

The ILO’s 2025 Working Paper 140 measures occupational exposure to generative AI by combining task-level data, expert input, and predictions from AI models. The study uses a revised framework of four progressively increasing exposure gradients, allowing researchers to distinguish degrees of possible impact instead of treating every occupation as equally vulnerable. 1

The global picture is significant but more limited than many headlines imply. The ILO reports that one in four workers worldwide are in an occupation with some generative-AI exposure. Yet 3.3% of global employment falls into the highest exposure category. The corresponding shares are 4.7% of female employment and 2.4% of male employment, illustrating that exposure is also connected to how work is distributed across occupations. 1

Clerical occupations continue to show the highest exposure levels in the 2025 assessment, while some highly digitized professional and technical occupations show increased exposure. These findings describe patterns across the global labor market, not a ready-made map of any individual publisher’s workforce.

Why task analysis is more useful than job labels

Most occupations contain a mixture of activities. Some may be more amenable to assistance from generative AI, while others depend on human judgment, context, accountability, or interaction. For that reason, the ILO’s 2025 study says transformation of jobs is the most likely impact, because most occupations still include tasks requiring human input. The ILO’s related 2023 global analysis likewise concluded that the overwhelming effect of generative AI would be to augment occupations rather than automate them. 2

For a newsroom, this suggests mapping work at a practical level. A team might review reporting, research, transcription, translation, headline development, audience analysis, production, editing, and administrative tasks separately. The purpose is not to assign a simple “safe” or “at risk” label. It is to ask what assistance could change the sequence of work, where review remains essential, and which responsibilities should stay clearly with people.

That exercise should also examine quality and risk. A task that can be accelerated is not automatically suitable for unrestricted automation. Editors may need to consider accuracy checks, attribution, confidentiality, originality, and the consequences of an error. These are newsroom-specific questions that cannot be answered by a global exposure percentage.

From estimates to a responsible transition plan

Once tasks are mapped, leaders can compare possible changes with the people who perform and oversee the work. Human consultation is important not only for fairness but also for operational accuracy: staff often understand exceptions, bottlenecks, and quality controls that do not appear in formal job descriptions. Local task analysis and validation are necessary before changing workflows, expectations, or role definitions.

Planning can then focus on several concrete choices: which tasks might be augmented, which processes need additional review, what training is required, and how changes will be evaluated over time. A pilot may reveal that a tool helps with one activity but creates new checking work elsewhere. That result is not evidence that a global forecast was wrong; it is evidence that implementation depends on local conditions.

The ILO’s research also links job quality and fair-transition policies to social dialogue and adequate regulation. In practice, that means treating adoption as an organizational change process rather than a software purchase. Newsrooms should document decisions, explain how responsibilities are shifting, provide channels for staff feedback, and revisit assumptions as tools and working methods evolve.

A better question for newsroom leaders

The most durable question is not “How many jobs will AI replace?” It is “Which parts of this work may change, and how can we manage that change responsibly?” The ILO evidence supports a measured answer: generative AI may affect many occupations, but the most likely outcome is transformation and augmentation across tasks, not automatic elimination of whole jobs. 1 2

That conclusion is neither a guarantee of job preservation nor a prediction for a particular employer. It is a planning discipline. By combining task-level review, staff consultation, human validation, and attention to job quality, a newsroom can respond to technological change without mistaking a global, model-based estimate for a local employment decision.