Computer and Mathematical · United States (national)
Computer and Information Research Scientists
How exposed this work is to automation, shown as ranges with the reasoning and sources.
The short version: this role is moderately exposed to automation, and a near-term shift looks unlikely.
It is a sourced range, not a verdict, and the detail is below.
Built on O*NET tasks, BLS pay and projections, and peer-reviewed AI-exposure research. Sources shown.
Data snapshot SNAPSHOT-2026-11-06. How it works
AI / softwarerange 38.0 to 62.0
Physical / roboticsrange 3.0 to 27.0
When the change is likely
Largely unaffectedlocal demand growing
What to do about it
Use AI to outpace AI.
Strategies to protect your livelihood, most important first.
1
Stay sharp; urgency is low
This role looks durable. Keep your skills current and watch for shifts in your field.
2
Lean into the human edge
The exposed parts are routine and cognitive. Invest in judgment, client relationships, and complex problem-solving that AI still cannot do.
3
Pivot toward Software Developers
It is skill-adjacent and less exposed (44.7 vs 50). A lateral move now beats a forced one later.
4
Build a financial buffer
Keep 6 to 12 months of expenses and diversify your income, so a disruption is a setback rather than a catastrophe.
5
Add a recognised credential
A certification or license raises the bar against automation and signals value in a tightening market.
Keep the monitor on
Free shows where you stand today. JobBreak Monitor writes your full AI-researched plan and keeps watching, so you know the moment your outlook shifts.
Free
$0
See exactly where you stand.
Full job-risk score & timeline
Task-by-task exposure breakdown
One starter strategy
AI-researched personal plan
Ongoing monitoring & alerts
Current plan
Keep ahead
JobBreak Monitor
$3.99/mo
$29/yr
or $29/yr save 39%
that's about $2.42/mo, billed yearly
Your full researched plan, plus a watch on your job.
Everything in Free
Full AI-researched, cited action plan
Resume-aware personalisation
Re-run your plan any time
Alerts when your outlook shifts
Start monitoring
Secure checkout by Stripe. Cancel anytime.
Which of my tasks are exposed
Analyze problems to develop solutions involving computer hardware and software.
50.0
Apply theoretical expertise and innovation to create or apply new technology, such as adapting principles for applying computers to new uses.
50.0
Assign or schedule tasks to meet work priorities and goals.
50.0
Meet with managers, vendors, and others to solicit cooperation and resolve problems.
100.0
Design computers and the software that runs them.
50.0
Conduct logical analyses of business, scientific, engineering, and other technical problems, formulating mathematical models of problems for solution by computers.
50.0
Evaluate project plans and proposals to assess feasibility issues.
50.0
Participate in multidisciplinary projects in areas such as virtual reality, human-computer interaction, or robotics.
50.0
Longer, orange bars are more automatable; short green bars are resistant.
What is driving my result
lowers risk
raises risk
Task: Meet with managers, vendors, and others to solicit cooperation and resolve problems.
Task: Analyze problems to develop solutions involving computer hardware and software.
Local demand is growing
Lower-risk adjacent roles
Software Developers
44.7
Computer Systems Analysts
47.1
A full, ranked mitigation plan is part of the paid tiers.
How this score is calculated
Map Computer and Information Research Scientists to its standard O*NET task list (8 core tasks).
Score each task for AI/software exposure (Eloundou et al. 2023) and physical/robotics exposure (an O*NET work-context proxy), each 0 to 100.
The two axis scores are the task averages: AI 50.0, robotics 15.0.
The timeline comes from the higher axis (50.0), shifted for local demand (growing), and seniority. It is read against your remaining working years.
Ranges are about ±12 points. A backtest against real BLS employment change shows the risk ranking holds in aggregate, but individual outcomes are predicted weakly, so results are ranges, never a flat verdict.
Sources: O*NET, Eloundou et al. (2023), GPTs are GPTs, O*NET Work Context, BLS CPI-U, BLS OEWS.
Snapshot SNAPSHOT-2026-11-06. Full method and backtest on How it works.