How we determine AI impact on your work

Most "AI job risk" content gives you predicted statistics or a list of jobs that might disappear. That is not what this is.

Remix Career reads AI impact at the task level: what parts of work like yours are likely to be automated, augmented, or anchored in human judgment, and what that means for how you position your experience next.

The free AI Impact Check applies this method to your own summary of job activities or your job description. The role library applies it to 25 common mid-career roles in depth.

This page documents only the free AI Impact Check. For how Remix Career builds your full Career Strategy from reflective answers, archetypes, and coaching methodology, see The Remix Career methodology.

What "exposure" means, and what it does not

Exposure here means how much of the typical task mix in a role is susceptible to AI-driven change, based on published occupational research and task profiles. It is not a prediction about whether you will lose your job.

What exposure means

• Which categories of tasks are shifting: drafting, analysis, coordination, judgment.

• Where AI is likely to take more of the keystroke layer.

• Where human judgment, relationships, and accountability still anchor the role.

• A fit-first read on where your experience still compounds.

What exposure does not mean

• A layoff forecast for your employer.

• A single score that ranks you against other people.

• Legal, financial, or HR advice.

• A guarantee about any specific tool or model.

We use bands, for example "high exposure on routine modeling and moderate on judgment", rather than precise percentages. Occupational data does not support precision at the individual level, and a band is more honest about what the evidence can support.

The Remix task-based method

We do not start from "Will AI replace this job title?" and work backward. We start from tasks, the repeatable units of work occupational researchers already catalog, and ask which tasks AI is most likely to absorb, accelerate, or leave alone.

Every "read" on this impact is organized into three lanes.

Automate

Tasks where AI can increasingly handle the full loop: gather inputs, produce a first draft, reconcile data, run standard templates. These tasks shrink as a share of a strong performer's week even when the role itself persists.

Augment

Tasks where AI speeds up the work but a human still owns the frame: scenario analysis, pattern detection, draft-and-refine cycles, faster prep for decisions that still require a person in the room.

Anchor

Tasks that stay human because they depend on accountability, trust, ambiguity, taste, or stakes: framing the right question, choosing assumptions, influencing stakeholders, signing off on a call.

The practical question is not whether your job is doomed. It is whether your week is still weighted toward work that can be automated, or whether your activities are more oriented around what is now considered anchor work.

When you use the AI Impact Check, we match your title to our role research library, then, when your description is specific enough, personalize the task lanes from what you actually said you do. Generic titles get library defaults. Specific descriptions get a sharper mirror.

How we assign exposure bands

Role-level bands in the library are built from a reconciled evidence stack, not a single index.

1. AIOE (AI Occupational Exposure). Felten, Raj and Seamans, A Method to Link Advances in Artificial Intelligence to Occupational Abilities (2021). Establishes which occupational abilities correlate with AI capability advances.

2. Task-level LLM exposure. Eloundou, Manning, Mishkin and Rock, GPTs are GPTs (2024). Estimates the share of work tasks exposed to large language models by occupation.

3. O*NET task profiles. U.S. Department of Labor. Ground-truth task lists and work activities by SOC occupation code.

4. Labor-market direction. U.S. Bureau of Labor Statistics, Occupational Outlook Handbook. Employment projections and role context.

5. Institutional synthesis. World Economic Forum Future of Jobs Report (2025 edition), plus selective use of Tier-2 institutional research where it sharpens task-level interpretation.

The band assignment rule, in plain language: we compare a role's relative position on AIOE and Eloundou-style task exposure, cross-check against the O*NET task mix, and reconcile with current labor-market signals. When sources disagree, we widen the band or split the read, for example high on routine deliverables and moderate on judgment, instead of forcing a single number.

We never invent a precise percentage of your job at risk, rank you against other users, or present model output as peer-reviewed fact.

Source tiers

Tier 1: peer-reviewed and government

AIOE, Eloundou et al., O*NET, the BLS Occupational Outlook Handbook, and OECD employment work. These are the primary inputs to exposure bands and task categorization.

Tier 2: institutional and industry

WEF Future of Jobs, the Anthropic Economic Index, and major consultancies. We use these for directional confirmation, the 2026 read context, and decline or growth narratives.

Tier 3: product and news

Vendor announcements and media trend pieces. These are never used to set bands. They may inform plain-language examples only.

The full bibliography with links is at the bottom of this page.

Personalization and library defaults

• Role title only, or a very thin description. You get the matched role profile from the 25-role library, or a category fallback.

• A specific description of tasks, tools, and decisions. You get personalized automate, augment, and anchor lanes, constrained by the matched role's evidence.

• Email unlock (optional). You get the full playbook, adjacent directions, and a 30-day experiment. This is still not a paid career assessment.

Personalization is constrained. The model cannot invent exposure bands or sources that contradict the library record. That keeps the free tool fast, citable, and consistent with the role pages.

Limitations

• Role evidence, not employer truth. Your company's stack, politics, and customers may accelerate or delay change.

• Point-in-time research. AI capability moves quickly. We update the library on a published cadence, not in real time.

• Mid-career framing. This method is built for people with enough experience to describe real work patterns, not for entry-level labor-market forecasting in the abstract.

• Not the full Remix Career assessment. The paid Role Strategy synthesizes your answers, resume, and coaching methodology into directions, listings, and skills guidance. The AI Impact Check is a free, task-level lead-in. The two are related but not interchangeable.

Update cadence

• Role library (25 roles). Reviewed at least semi-annually, and ad hoc when major labor-market releases land such as the BLS Occupational Outlook Handbook or the WEF Future of Jobs report.

• AI Impact Check tool copy and matchers. Updated with library releases.

• This methodology page. Updated when band rules or primary sources change. The last reviewed date is shown at the foot of this page.

Library version: 25 roles, deepened standard, June 2026 evidence pass.

Fit-first interpretation

Remix Career is built on a coaching premise: the useful move is usually repositioning toward anchor work and adjacent roles, not panic about automation headlines.

That is why every role read includes a central shift line describing what changes in the task mix, adjacent directions that reuse the same underlying strengths, and a practical next step you can run this week rather than a default instruction to learn to code.

If the read surfaces high automate exposure, that is an invitation to migrate your week toward augment and anchor tasks, and to explore directions where your judgment is the product.

Frequently asked questions

Why bands instead of percentages?

Because occupational exposure estimates are noisy at the individual level and vary by source. Bands communicate uncertainty honestly and are easier to act on.

Is this predicting replacement?

No. It maps how the task mix in work like yours is likely to shift. Many roles persist while the work inside them changes.

How is this different from the AI Impact Analysis in the paid report?

The free Check is task-level and library-backed. The paid analysis sits inside your full Role Strategy and connects AI impact to your recommended directions, skills, and listings.

Can I cite this page?

Yes. Attribute to Remix Career and link to this page.

Primary sources

Use these for citation. Where you need a single reference, prefer linking to this page as the methodology summary.

1. Felten, E., Raj, M. and Seamans, R. (2021). A Method to Link Advances in Artificial Intelligence to Occupational Abilities. AIOE scores and methodology. aioe.org

2. Eloundou, T., Manning, S., Mishkin, P. and Rock, D. (2024). GPTs are GPTs: Labor market impact potential of LLMs. openai.com/research/gpts-are-gpts

3. O*NET OnLine. U.S. Department of Labor, Employment and Training Administration. Occupational tasks, skills, and work activities. onetonline.org

4. U.S. Bureau of Labor Statistics. Occupational Outlook Handbook. Employment projections and role descriptions. bls.gov/ooh

5. World Economic Forum (2025). The Future of Jobs Report 2025. weforum.org

6. Anthropic Economic Index. Observed AI use patterns by task category, used as supplementary Tier 2 evidence. anthropic.com/economic-index

Keep going

Run the free AI Impact Check on your own role, browse the AI impact library by role, or read how we build a full Role Strategy.

Last reviewed: July 2026