The task-level exposure audit (do this, it takes an hour)
List your weekly tasks, score them without flinching, and find out what actually survives automation.
Most people already know their job is at risk. They just haven’t said it out loud yet. There’s a specific kind of knowing that sits below the surface of daily routine. You feel it in the meeting where the AI demo does your Thursday in forty minutes. You feel it when the graduate scheme gets paused for the second year running. You feel it in the silence after your manager says “we’re looking at efficiencies,” a phrase that has never once meant anything good for anyone in the room.
Knowing and auditing are different things. Knowing is a feeling. An audit is a number. And the distance between the two is where most people hide.
This chapter is about closing that gap. Not with false comfort, not with panic, but with honest arithmetic.
Your job is not one thing. It’s a bundle of tasks, and each task has a different exposure level. This is the most important shift in the whole audit. Nobody’s role is entirely safe or entirely doomed. It’s a mix, and the proportions matter.
Take a marketing manager. On a given day, they might pull data from analytics platforms and build a report. High exposure, pattern-based, AI does this now. Write a strategy brief from a template. High exposure, documentation-heavy, AI generates these. Sit in a meeting with a client and read the room when the conversation gets tense. Low exposure, requires emotional context and relationship history. Manage three agency relationships through email coordination. High exposure, pure orchestration. Make a judgment call about whether to shift budget mid-campaign based on instinct and data. Moderate exposure, judgment component matters, but AI-augmented decisions are coming.
That’s five tasks. Three are highly exposed. One is moderately exposed. One is relatively safe. If 60% of your day is spent on the high-exposure tasks, your role is in trouble even if the remaining 40% is distinctly human. Because your employer is paying you for the full day, and if three-fifths of that day can be done by software at a fraction of the cost, the maths stops working in your favour.
The question isn’t whether your job disappears. It’s whether enough of your job disappears that the remaining tasks don’t justify your salary.
Here’s how to do the task-level audit. Take a piece of paper and a typical week. List everything you do. Mark each task: high, moderate, or low exposure. Be honest. Nobody else is going to see this list. If you catch yourself writing “strategic leadership” when what you actually do is chair a weekly standup and forward the notes to your manager, fix it. The audit only works if it’s honest.
A business analyst I know ran this exercise over a weekend. Her weekly hours broke down roughly like this: twelve hours on data extraction and report building, high exposure. Six hours on stakeholder meetings and relationship management, low exposure. Eight hours on email coordination between departments, high exposure. Five hours on genuine strategic thinking, connecting patterns the data revealed to business decisions nobody else had noticed, low exposure. And nine hours on documentation, meeting notes, status updates, process paperwork, very high exposure.
That’s twenty-nine hours of high or very high exposure out of forty. Seventy-two per cent. The eleven hours of lower-exposure work, the relationship management and the strategic thinking, were real. They were valuable. They were also not enough to justify a full-time salary when the other twenty-nine hours could be done by a language model and an automation workflow.
She didn’t like the number. Nobody does. But honest numbers are the ones you can plan around.
One thing the task-level audit misses: velocity. Your exposure score today is a snapshot. The more important question is how fast it’s changing. AI capability is not advancing linearly. It’s compounding. The tools available in 2024 handled basic reporting and content generation. The tools available in 2026 handle multi-step reasoning, cross-platform coordination, relationship tracking, and contextual judgment. The gap between “AI can theoretically do this task” and “AI does this task better and cheaper than a human” is closing faster than most people’s career plans account for.
A role that scores 2.5 on task-level exposure today might score 3.5 by next year. Not because the role changed. Because the tools improved. The audit needs to be dynamic, revisited quarterly, adjusted as capabilities evolve. A score you’re comfortable with in March might be uncomfortable by September.
This isn’t meant to create panic. It’s meant to create urgency. The difference between the two is that panic produces poor decisions and urgency produces timely ones.
Now, the household audit. Individual risk is one thing. Household risk is what actually matters, because bills don’t care whose income pays them. Most discussions about career risk focus on the individual. But people don’t live as individuals. They live in households. Shared expenses, shared debts, shared dependants, shared assumptions about how the mortgage gets paid.
Ask uncomfortable questions. Are both earners in high-exposure sectors? If you’re a marketing manager married to a business analyst, you’ve got correlated risk. Both roles share the same exposure profile: orchestration, text-based output, professional services. If your employer automates your role and your partner’s employer does the same thing six months later, you’re not dealing with a career setback. You’re dealing with a household crisis.
What’s your financial runway? The FCA surveys are consistent: one in four UK adults has low financial resilience. A third have less than five hundred pounds in emergency savings. Eleven million Britons have less than a thousand pounds to draw on. If your household is in these categories and both incomes are exposed, the audit produces a number that should alarm you. Not because alarm is useful on its own. Because it should accelerate the timeline on the strategies that come next.
The average UK adult under 55 has about ten thousand pounds in savings. For most households, that’s two to three months of expenses. If your role disappears and you’re job-hunting for six months, the savings don’t cover it.
What are your fixed costs? Mortgage or rent, childcare, school fees, car finance, insurance, utilities. These don’t flex when income drops. A household spending 40% or more of income on housing, which describes a large chunk of UK households in major cities, has almost no margin for income disruption. The gap between “uncomfortable” and “crisis” is about three months.
Is there a fallback? Family resources, geographic flexibility, skills that transfer to lower-risk sectors. Not nice-to-haves. The difference between a managed transition and a panic.
Dual incomes create the illusion of resilience. The reality depends on sector diversification. Two incomes are better than one. Obvious. But two incomes in the same sector, or in sectors with correlated risk profiles, create a different kind of vulnerability. Not twice as safe. Twice as exposed to the same shock.
A household where one partner works in healthcare and the other in digital marketing has genuine diversification. If the marketing role gets automated, the healthcare income continues. The household absorbs the shock. A household where both partners work in professional services, one in consulting and one in accounting, has concentration risk. The same AI adoption wave that restructures the consulting firm will restructure the accounting firm. Not necessarily on the same day. But within the same window, driven by the same competitive pressure, producing the same result: fewer roles, lower pay, or both.
The audit question for dual-income households is simple: write down both roles, assess both for task-level exposure, and ask whether the risks are correlated. If they are, your household risk is significantly higher than either individual risk suggests. And the practical follow-through: if the risks are correlated, one partner should be actively positioning toward a lower-risk sector. Not both simultaneously. That creates chaos. But a deliberate, planned diversification where one household income moves to a less exposed area while the other maintains the current trajectory. Household-level portfolio management. It’s as important as anything in a pension statement.
The skills half-life problem sits underneath all of this. The skills that got you here may not get you through what’s coming. And the speed at which skills depreciate is accelerating. Skills half-life, the time it takes for half of what you know to become outdated, is now estimated at under two and a half years in technical fields. Half of what a professional learned in a training course eighteen months ago is already losing relevance.
For the honest audit, this means something uncomfortable: the skills on your CV, the ones you list under Core Competencies or Areas of Expertise, may already be past their peak market value. Not because you’re bad at them. Because the market’s need for a human to perform them is declining.
The audit should include a skills currency check. When did you last learn something actually new, not a refresher, not an update, but a capability you didn’t have before? If the answer is more than two years ago, your skills portfolio is depreciating faster than you’re replenishing it. That’s a position you can change. But you need the honest assessment first.
The honest audit doesn’t tell you what to do. That’s the next part. It tells you where you stand. And where you stand determines what’s available to you.