Technology,
automation, and due process in unemployment insurance

State unemployment systems use technology for document processing,
identity verification, fraud detection, task routing, fact-finding,
scheduling, records, and other functions. The U.S. Department of Labor
has published resources on automation in UI workflows and research
concerning artificial-intelligence assistance for adjudicators.

Worker comparing an unemployment document with information on a laptop.

Technology may reduce delay and improve consistency. It may also
create problems when a classification, identity determination, fraud
flag, notice, or summary is inaccurate, opaque, or difficult to
challenge.

The claimant-side questions

When technology may have affected the case, ask:

  • What decision was actually made, and by whom?
  • What system, data, or rule produced the disputed output?
  • Is the output a record, summary, prediction, recommendation, or
    final determination?
  • Was the information attributed to the correct claimant and benefit
    period?
  • Did a human decision-maker review the material and exercise required
    judgment?
  • What correction or appeal process exists?
  • What logs, notices, source records, or explanations are
    available?
  • Did a portal or authentication failure interfere with filing or
    participation?

The right to obtain particular records or explanations varies by
jurisdiction.

Evidence and reliability

A computer-generated record is not self-explanatory. Its weight may
depend on the source data, completeness, system function, attribution,
error rate, human verification, and the ability to test the conclusion.
A risk flag or summary is not necessarily proof of the underlying
fact.

Access and equity

Portal design, authentication, language access, disability access,
device compatibility, and notice clarity can determine whether a
claimant files, participates, or preserves an appeal. Document technical
problems with dates, screenshots, error messages, account-reset history,
calls, emails, and prompt follow-up.

Labor-market context

AI and robotics are changing hiring, job design, and workforce
demand. Those changes may affect separation disputes, job-search
requirements, suitable-work questions, benefit administration, and the
volume and complexity of unemployment claims. Legal analysis should
distinguish observed evidence from forecasts and assumptions.

The due-process objective

The point is not to oppose modernization categorically. Efficiency
should not replace fair notice, reliable fact development, human
accountability, meaningful hearing rights, or a decision that can be
reviewed on an adequate record.