Durable Earnings

Building income that lasts in a world that’s changing fast.

The 5-Point AI Vulnerability Assessment for Your Role

—

by

Neil Gulino
in

If you’re over 40 and trying to figure out what AI means for your job, the worst place to start is your title. “Project manager,” “analyst,” “operations lead,” “marketing director” – those labels are too blunt to tell you much. AI doesn’t replace titles first. It strips out tasks, compresses teams, and makes companies wonder how many people they still need once the boring parts move into software.

That’s why job security now has a costume. It still looks like a normal paycheck, a normal org chart, and a normal Monday. Underneath, the work may already be changing.

A real AI vulnerability assessment for your role isn’t about panic. It’s pattern recognition. The goal is to figure out whether your value comes from repeatable output, judgment under pressure, cross-functional usefulness, or some mix of the three. Once that’s clear, you can make better decisions about where your career stands.

Point 1: Score Your Task Composition – How Much of Your Day Is Automatable?

Start with the most uncomfortable question: what percentage of your week is made up of repeatable tasks that follow known rules?

That includes things like summarizing meetings, drafting standard emails, preparing recurring reports, updating CRM fields, scheduling, data cleanup, formatting decks, and pulling the same numbers for the same audience every Friday because apparently ritual suffering is still a management system. If a task can be described in a checklist with eight predictable steps, AI is already circling it.

The numbers are not subtle. OpenAI, the University of Pennsylvania, and OpenResearch found in 2023 that about 80% of the U.S. workforce could see at least 10% of their tasks affected by GPTs, and roughly 19% of workers could see at least 50% of their tasks exposed. Their test was straightforward: could an LLM cut the time for a task in half without hurting quality? Using O*NET data across 1,016 occupations, the answer was yes for a lot of office work. Goldman Sachs later found office and administrative support roles had the highest automation exposure at 46% (OpenAI, University of Pennsylvania, and OpenResearch, 2023; Goldman Sachs, 2025).

That doesn’t mean every exposed job disappears. It means task-heavy roles get squeezed first. One person starts doing what two people used to do. A manager decides the team can “streamline.” Nobody says replacement. They say efficiency.

So score your day honestly:

  • Low vulnerability: less than 25% of your work is repetitive, rules-based, and easy to document
  • Moderate vulnerability: 25% to 50% fits that description
  • High vulnerability: more than half your value comes from predictable output

Be strict. If the task depends mostly on consistency, speed, formatting, retrieval, or standard wording, it belongs in the automatable bucket. If that’s most of your day, your role has exposure even if the title sounds senior.

Point 2: Map Your Skill Half-Life – Are Your Core Competencies Depreciating?

Some skills disappear. More often, they just lose market value faster than people notice.

That’s the real problem with AI exposure. Your job can survive while your bargaining power erodes underneath it. You still work. You still get paid. But the company no longer has to pay as much for the same output because software now handles the first draft, the first analysis, the first sort, or the first pass.

PwC’s June 2025 Global AI Jobs Barometer found wage premiums for AI skills reached 56%, up from 25% a year earlier. It also found skill change in AI-exposed jobs is 66% faster than in other occupations, and jobs with high AI exposure are growing 3.5 times faster than the overall market. At the same time, Pew Research reported in February 2025 that 52% of U.S. workers are worried about AI’s workplace impact, while only 6% think it will create more job opportunities for them personally (PwC, 2025; Pew Research Center, 2025).

That gap matters. Employers are paying more for people who can work with AI, redesign processes around it, or add human value after AI does the cheap part. They’re paying less for people whose main advantage was producing the first version of something.

Think about your own core skills and ask:

  • Are these skills getting more valuable, or just easier to imitate?
  • Would an employer pay a premium for them in two years?
  • Do they depend on software staying worse than it is now?

If your best skills are accuracy, speed, basic analysis, standard communication, or process compliance, the half-life may be shorter than you want. If your skills involve negotiation, relationship trust, conflict navigation, pattern judgment, or translating across departments, they tend to hold value longer.

A role can stay employed while its skills depreciate. That’s how people end up surprised by a 12% raise that never comes, a promotion that quietly goes away, or a layoff that looks sudden only because they missed the five smaller warnings before it.

Point 3: Audit Your Decision Authority – Can AI Replicate Your Judgment?

This is where a lot of people get clarity. Two jobs can look equally “knowledge based” from the outside and have very different AI vulnerability depending on what kind of judgment they require.

Tufts University’s AI Jobs Risk Index analyzed nearly 800 occupations and found project management specialists scored 87.5 out of 100 in AI exposure. Nursing assistants scored 18.2. Surgical assistants scored 7.1. A Forbes analysis in January 2026, using the Resume Now AI-Resistant Careers Index, found roles requiring real-time judgment under pressure and real accountability – like nurse anesthetists and emergency physicians – had some of the lowest automation risk (Tufts University Digital Planet, 2024; Forbes, 2026).

The distinction isn’t education level. It’s decision context.

If your role depends on handling messy variables, reading people, taking responsibility for consequences, and adjusting in real time when conditions change, AI has a harder time replacing the center of your value. If your role depends on making routine decisions inside a narrow framework, AI has a much easier path.

Ask yourself three questions:

  1. Are the stakes real? If your judgment is wrong, does something meaningful happen beyond a typo or delay?
  2. Are the variables unstable? Do you work in situations where the facts shift, people react unpredictably, or context matters more than rules?
  3. Are you accountable? Does someone trust you to own the decision, not just recommend an option?

High decision authority usually lowers vulnerability. Not because AI can’t generate answers, but because generated answers are not the same as owned judgment. Software can suggest. It can’t absorb blame, calm an angry client, manage political fallout, or make a call with incomplete information and live consequences.

If your role mostly routes decisions instead of owning them, that is a warning sign.

Point 4: Calculate Your Cross-Functional Breadth – Are You a Specialist or a Multi-Domain Operator?

Narrow expertise can still pay well. It just breaks faster when the narrow thing becomes cheaper.

IBM found in 2023 that executives estimate 40% of their workforce will need to reskill within three years due to AI and automation. McKinsey’s 2025 State of AI Survey found 88% of organizations now use AI in at least one business function, up from 78% the prior year, and generative AI adoption nearly tripled to 79%. PwC also found wages rise twice as fast in AI-exposed industries and revenue per employee grows three times faster – but those gains favor workers who develop skills AI doesn’t handle well, including persuasion, negotiation, emotional intelligence, and contextual judgment across domains (IBM, 2023; McKinsey, 2025; PwC, 2025).

Here’s the blunt version: if your expertise only works inside one lane, your risk is higher than you think.

Cross-functional breadth means you can connect operations to finance, customer needs to product decisions, data to execution, or strategy to the messy work of getting other humans to cooperate. Those bridges matter because companies don’t just need output. They need people who can move ideas across silos without turning every meeting into a hostage situation.

A useful way to score this point, especially if you’ve read Your Income in the AI Era:

  • Low vulnerability: you regularly solve problems across multiple functions or stakeholder groups
  • Moderate vulnerability: you work mainly in one function but often translate or coordinate with others
  • High vulnerability: your value depends on one narrow specialty with little crossover application

This is also where How to Use AI Tools to Work Faster Without Losing Your Job to Them becomes relevant. The safest position is not “person who avoids AI.” It is “person who uses AI for the mechanical stuff, then adds cross-functional judgment the tool can’t fake.”

Point 5: Locate Your Value Chain Position – Where Does AI Threaten Your Organizational Layer?

Not every layer of a company gets squeezed the same way.

Goldman Sachs estimates generative AI could expose the equivalent of 300 million full-time jobs globally to automation and could automate tasks that account for 25% of all U.S. work hours. Tufts projects 18.3% job loss in the Information sector and 16.5% in Finance and Insurance, while roles involving physical labor in unpredictable environments face less than 1% displacement risk. The World Economic Forum’s Future of Jobs Report 2025 projects data entry clerks declining 34% and administrative assistants facing significant reductions by 2030 (Goldman Sachs, 2025; Tufts University Digital Planet, 2024; World Economic Forum, 2025).

That means your vulnerability isn’t just about what you do. It’s also about where your work sits in the company’s value chain.

Roles in the middle layers are often the most exposed: reporting, coordination, documentation, routine analysis, internal communications, scheduling, compliance prep, and support work that exists to move information from one place to another. Those roles matter. They’re also exactly where software creates the strongest financial temptation.

By contrast, roles closer to revenue, trust, crisis handling, physical execution, or high-accountability decision making tend to have more insulation. Not immunity. Insulation.

If you’re not sure where you sit, ask this:

  • Does your work directly protect revenue, create revenue, or prevent expensive mistakes?
  • Or does it mostly package, relay, track, and standardize work created elsewhere?

That answer tells you plenty. So do Which Jobs AI Is Replacing First – and Which Ones It Isn’t and White-Collar Jobs Most at Risk from AI: A Plain-English Breakdown. Both make the point: exposure shows up first where work is abstract, repeatable, and easy to cost-cut.

FAQ

How often should I reassess my role’s AI vulnerability – is this a one-time exercise?

No. Reassess every six to 12 months, or sooner if your company rolls out new AI tools, restructures teams, or changes performance expectations. AI risk is not static because the software keeps improving and employers keep getting more comfortable redesigning work around it.

If my vulnerability score is high, should I start looking for a new job immediately?

Not automatically. A high score means you need options, not a panic move. Start by reducing your exposure inside your current role: take on work with more judgment, more cross-functional value, and more visible business impact. At the same time, update your resume, rebuild your network, and look at adjacent roles where your experience travels better.

What’s the single most protective skill I can develop to lower my vulnerability score?

Contextual judgment is the strongest broad protection. That includes deciding what matters, reading tradeoffs, handling ambiguity, and making calls other people trust. It’s not a single software skill. It’s the ability to combine facts, incentives, people, and consequences into a useful decision.

Can AI partially automate my role without eliminating my job entirely?

Absolutely. In fact, that’s the more common pattern. Partial automation usually shows up first as higher workload expectations, smaller teams, and pressure to produce more with the same headcount. The role survives, but the advantage shifts. That’s why partial automation still matters for pay and job security.

Does having a college degree or advanced certification help protect against AI displacement?

Sometimes, but not by itself. Credentials help when they signal judgment, regulation, accountability, or domain depth that software can’t easily replace. They help less when the job’s daily work is still repetitive and rules-based. Degree status is a weak signal compared with task composition, decision authority, and business context.

After running your own vulnerability assessment, one practical step is knowing where you stand financially – your credit picture is a baseline you’ll want in good shape if disruption affects your income. Credit Karma gives you free access to your score and alerts without selling you anything you didn’t ask for.

In the end, an AI vulnerability assessment is really a market reality check. It tells you whether your role is paid for judgment, breadth, and accountability, or mostly for output that can be copied cheaper.

If the score comes back high, that is not a verdict on your worth. It’s a signal to reposition early, while you still have room to move. Start there.

Affiliate disclosure: This article contains affiliate links. If you use them, Durable Earnings may earn a commission at no extra cost to you.

Sources

Continue reading: Read the pillar — Your Income in the AI Era

This article is for informational purposes only and is not financial advice. Consult a qualified professional for personalized guidance.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *