Your role may not exist in three years

Which roles in international organizations AI will change and what that means for you

The reports, the analyses, the documentation, the coordination, the translation, and the work that has defined your week, your grade level and your professional identity for years, is precisely the kind of structured, process-driven, repeatable work that AI absorbs first.

Your competence and your organization's regard for you are beside the point. The nature of the work itself is what has changed.

The United Nations system moves slowly, and we have all seen it take years to adopt technologies that the private sector had integrated and moved on from long before. Until recently, that institutional pace felt like a reasonable form of protection. Budget pressure is real, and the logic of AI-enabled cost reduction does not depend on how fast the organization moves, rather it comes down to a simpler question: can the same work be done with fewer people?

Disclaimer: I have no insider information about plans for AI integration in the UN system and what follows is not a prediction about what might happen. It is an assessment of what is already in motion, what it means for all professionals at P3 through D2 level inside the UN system, and what you need to do about it now.

Why the UN system feels protected and why that feeling is wrong

International organizations operate largely on majority decision-making, processes, and institutional memory, and historically the adoption of new technology has moved at snail’s pace. Having worked at the UN during the transition from electric typewriters to WANG computers, I can say that with confidence.

As a result, many professionals inside the system have built their sense of job security on that reality, and for a long time that was a reasonable position to hold.

Without exception however, every professional I have spoken to since the beginning of 2026 has told me the same thing, namely that they do not know to what extent their role will change when AI is integrated into their organization. That uncertainty is not a comfortable place to sit, and it is not one you can afford to stay in.

The roles that are changing

You must be able to look past your job title and examine your tasks in detail because that is where the real picture emerges.

Translation and interpretation

The UN operates one of the largest language services in the world; six official languages, enormous documentation volumes, and simultaneous interpretation across meetings at every level. AI translation is already being used inside several UN entities and the limitations are real because machine translation loses nuance, political tone and the interpretive judgement that makes a translated resolution represent what the drafters intended. 

Other high-volume, lower-stakes work such as routine circulars, standard administrative communications, and operational field reports, is increasingly handled by machine translation with light human review. 

Human oversight remains essential, but the number of people required has changed because the transactional volume has shrunk. The value has shifted entirely toward interpretation that carries diplomatic weight, cultural consequence and political precision, and a language professional whose value was primarily speed and volume is now in a fundamentally different position to one whose value is interpretive authority. If you cannot immediately distinguish which one you are, that is the first thing to resolve.

Programme administration and reporting

Programme administration includes a steady flow of progress and situation reports, results-based management documentation, and donor reporting. It follows a predictable structure, draws on standardised data, and produces outputs with defined formats, which are exactly the conditions in which AI performs well. The likely outcome here is that your role shrinks to the 15 per cent of it that requires genuine judgement, and the organization structures accordingly.

The programme officer whose primary contribution it is to transform field data into donor-ready reports is more exposed than the one whose contribution is knowing which field reality the data is failing to capture and being trusted enough to say so. The question worth considering is, which one are you, and can you demonstrate it when it matters.

Data analysis, monitoring, and evaluation

Monitoring and evaluation roles centred on data compilation, indicator tracking and standard analysis are changing faster than most professionals inside those roles might want to admit.

AI systems can now process large volumes of field data, flag anomalies and generate draft analytical summaries faster and more consistently than a mid-level analyst working alone. What remains human is interpretation, namely what the data means in context, what the organization cannot conclude from data alone, and what it should do next. Those things are very specific, and they require a different kind of professional confidence than simply compiling the data.

Human resources, recruitment, and administrative services

Human resource functions, for example vacancy management, job classification, and entitlement processing, follow rules-based processes at high volume. IBM's internal AI system now handles millions of HR interactions annually, resolving the vast majority without human escalation. UN HR functions are not immune to this logic and the budget pressure argument is already being made inside several UN entities.

The administrative roles that will survive are the ones that manage exceptions, exercise discretion, and hold institutional relationships, not the ones whose primary function it is to process transactions.

Procurement and contract management

Standard procurement processes such as solicitations, evaluation scoring against pre-defined criteria, and contract compliance tracking, are high-volume, rules-based and therefore well-suited to AI systems. Legal review, strategic supplier decisions and politically sensitive procurement will remain human. The administrative infrastructure surrounding them may not.

Where the human is still essential

The functions that remain firmly human are those that carry political, ethical, or legal consequence, such as relationship management with member states, negotiation under ambiguity, leadership in a crisis, cultural and contextual interpretation that no training dataset has yet captured, and the management of people through uncertainty.

No AI system has yet demonstrated the capacity to replace any of them.

The question every professional must answer however is whether your own work actually contains those functions, not the tasks and responsibilities listed in your job description, but in the work you did this week. If the honest answer is that it does not, that is important information, and it deserves a serious response.

The opening that IT professionals must act on now

IT professionals inside international organizations have spent the last decade managing legacy infrastructure, supporting helpdesks and maintaining systems built before cloud architecture was standard. The headline that AI will replace programmers however is not necessarily true.

What AI has done is to reduce the cost and time involved in writing code, which means that an IT professional who no longer spends most of their time on development tasks need not be redundant, because they are available for something considerably more valuable.

The more important question though is whether they will move fast enough to claim it, because the real opportunity is AI implementation. The vast majority of organizations are now using or exploring AI, yet successful implementation at scale remains the exception rather than the rule.

The gap between the intention to adopt and implement successfully is enormous, and the professionals who can bridge it - those who understand the technology and the institutional context together - are in short supply.

Organizations without in-house AI expertise face vendor dependence, slower adoption and increasing strategic risk during a transition period that will not pause whilst they catch up. That expertise gap represents one of the most significant career opportunities available to IT professionals right now.

And then there's the question of intellectual property

The intellectual property (IP) dimension also belongs here because inside organizations deploying AI systems, it is fundamentally an implementation problem. Every piece of content that an AI system generates or acts upon carries intellectual property implications; procurement documents, policy frameworks, training data drawn from institutional archives.

None of this is legally neutral and WIPO has been clear that the intersection of AI and IP governance is one of the defining legal frontiers of this decade, and inside the UN system, where the organization's own outputs interact with member state documents, third-party data and confidential field reporting, the exposure is significant.

The IT professional who understands AI implementation and has working knowledge of how IP intersects with AI deployment will be an essential part of the infrastructure for any organization serious about deploying AI responsibly, and that combination is currently rare enough to constitute a real competitive advantage.

The action you must take now

Go through your actual tasks from the last four weeks, not your job description and not what you aspire to contribute in your best moments, but what you actually did this month or this quarter.

Write them down and for each one, ask a single honest question: could a well-configured AI system produce a usable first draft, or handle this task entirely, now or within the near future?

The tasks that survive that exercise are the foundation of your next positioning, the capabilities you develop further, the contribution you make visible, and the professional identity you build. The tasks that do not survive are the ones to stop treating as the core of what you offer.

Your colleagues are not doing this exercise and it will be the professionals who take the time to answer the above, and then act on what they find, who will look back on this period as the moment they consolidated their value rather than allowed it to slowly erode alongside everyone else who waited.

The standard

Waiting for the UN to issue formal guidance on AI before you assess your own position is handing control of your career to an institution that has historically taken years to reach consensus on the adoption of new technology.

The standard required now is personal. It does not demand that you predict the future with precision but it does require you to audit your value, what is already changing, and what your next deliberate move needs to be.

The professionals who will lead in this environment are the ones who looked at the same information available to everyone, applied it to their own situation with honesty, and moved.

If you have read this and you are not certain which of your tasks will survive the next two years, that uncertainty is the starting point.

The Exposure Audit is a 90-minute session designed for exactly this moment, a precise assessment of where your value sits, what is already at risk, and what your next move needs to be. No generalisations, no generic frameworks, and certainly no advice that could apply to anyone. A clear-eyed picture of your specific position and a 30-day action plan built around it.

Book your Exposure Audit here: The Exposure Audit | GwynethEL - Feelgood