Explore the rankings
Legal · Updated Aug 2026

Paralegals and Legal Assistants

Assist lawyers by investigating facts, preparing legal documents, or researching legal precedent. Conduct research to support a legal proceeding, to formulate a defense, or to initiate legal action.

JVS 2.0.0-phase4b
AI Exposure
73/100
High

How much of this occupation’s work can be materially affected by current AI systems.

Replacement Risk
61/100
High

How likely exposure is to translate into reduced human demand.

Includes provisional estimates for AI adoption pressure and labour-market resilience. How this is measured

Evidence quality
Confidence83/100
Task coverage87%

Confidence reflects task coverage, mapping and capability-evidence quality, and how much of the score rests on provisional inputs.

Task-level evidence

What is driving the score?

Occupation scores are built from the task mix—not a single prediction about a job title.

JVS 2.0.0-phase4b
TaskImportanceAI impactExposure
Prepare affidavits or other documents, such as legal correspondence, and organize and maintain documents in paper or electronic filing system.High
74
Prepare, edit, or review legal documents, including legislation, briefs, pleadings, appeals, wills, contracts, and real estate closing statements.High
75
Investigate facts and law of cases and search pertinent sources, such as public records and internet sources, to determine causes of action and to prepare cases.High
75
Gather and analyze research data, such as statutes, decisions, and legal articles, codes, and documents.Medium
74
Meet with clients and other professionals to discuss details of cases.Medium
63
Prepare for trial by performing tasks such as organizing exhibits.Medium
76
Direct and coordinate law office activity, including delivery of subpoenas.Medium
74
Arbitrate disputes between parties and assist in the real estate closing process, such as by reviewing title searches.Medium
75
Appraise and inventory real and personal property for estate planning.Medium
74
Most exposed

Where AI can do more

Routine, digitized, and highly repeatable tasks face the greatest pressure.

  1. Prepare for trial by performing tasks such as organizing exhibits.76
  2. Prepare, edit, or review legal documents, including legislation, briefs, pleadings, appeals, wills, contracts, and real estate closing statements.75
  3. Investigate facts and law of cases and search pertinent sources, such as public records and internet sources, to determine causes of action and to prepare cases.75
  4. Arbitrate disputes between parties and assist in the real estate closing process, such as by reviewing title searches.75
Hardest to automate

Where people still matter

These tasks score lowest on automation feasibility—physical presence, judgement, accountability and real-world variability all resist end-to-end automation.

  1. Prepare affidavits or other documents, such as legal correspondence, and organize and maintain documents in paper or electronic filing system.01
  2. Prepare, edit, or review legal documents, including legislation, briefs, pleadings, appeals, wills, contracts, and real estate closing statements.02
  3. Investigate facts and law of cases and search pertinent sources, such as public records and internet sources, to determine causes of action and to prepare cases.03
  4. Gather and analyze research data, such as statutes, decisions, and legal articles, codes, and documents.04
  5. Prepare for trial by performing tasks such as organizing exhibits.05
Where else this work leads

Related occupations

Occupations O*NET links to this one. Relatedness reflects shared work, not a claim that these roles are safer.

See all rankings →
Beyond AI capability

Adoption and labour-market outlook

Structural factors are kept separate from raw capability so you can see what actually resists automation. Adoption pressure and labour-market resilience are still provisional models—25% of this occupation’s replacement-risk weight rests on them.

Human dependency70
Physical dependency22
Adoption pressure53
Labour-market resilience53
Methodology & sources

O*NET 30.3 occupational data interpreted through the JobsVsAI capability, automation and structural-constraint models.

Confidence83/100
CalculatedAug 21, 2026
Read methodology →