Hiring & Non-Backfills
Evidence-based tracking of reduced hiring, non-backfills, attrition and other ways AI can reduce demand for human labor without layoffs.
Which jobs are disappearing without layoffs?
The Hiring & Non-Backfills Tracker documents evidence that artificial intelligence is reducing demand for human labor through lower hiring, positions that are not replaced when employees leave, and workforce reductions accomplished through normal attrition.
These changes can be largely invisible in conventional layoff statistics. An employee may resign and never be replaced. A company may reduce planned hiring because AI allows existing employees to produce more. An entry-level position may simply never be created.
No layoff is required for human labor demand to decline.
Our standard is evidence. We look for company disclosures, executive statements, hiring data, administrative employment records and other evidence connecting AI adoption or AI-driven productivity to reduced hiring or nonreplacement.
Direct AI-attributed layoffs and other explicit workforce reductions are tracked separately in AI Job Displacement.
DOCUMENTED HIRING & NON-BACKFILL EVIDENCE
This tracker records evidence that AI is affecting employment through reduced
hiring, nonreplacement, attrition and other changes that may occur without
traditional layoffs.
LABOR-MARKET EVIDENCE
STANFORD DIGITAL ECONOMY LAB
AUGUST 2026
Employment weakness among young workers in AI-exposed occupations operates
primarily through reduced hiring rather than increased separations.
Finding
Using administrative payroll data covering millions of workers,
researchers found that employment among workers ages 22–25 in highly
AI-exposed occupations had fallen substantially relative to less-exposed
workers. The researchers found that the decline operated primarily
through reduced hiring of young workers rather than increased separations.
Signal
Reduced entry-level hiring
Evidence strength
DOCUMENTED
Why it matters
This provides administrative labor-market evidence for a mechanism that
conventional layoff statistics can miss: employment can decline because
companies hire fewer workers, even when existing employees are not being
dismissed.
U.S. CENSUS BUREAU RESEARCH
APRIL 2026
Early-career hiring declined sharply at employers with greater exposure to AI.
Finding
Analysis of matched employer-employee administrative data found an
immediate, sizable and persistent decline in hiring of workers ages
22–24 following the introduction of ChatGPT in the most AI-exposed
industry-state groups. The decline in hires was identified as the
primary cause of the subsequent employment decline.
Signal
Reduced early-career hiring
Evidence strength
DOCUMENTED
Why it matters
The findings independently reinforce the possibility that an important
early employment effect of AI is appearing in hiring flows rather than
mass layoffs.
COMPANY EVIDENCE
UBER
2026
AI productivity gains allowed Uber to moderate the pace of hiring.
Mechanism
Reduced pace of hiring
Evidence
Uber reported widespread adoption of AI coding tools among its engineers
and said productivity gains from AI allowed the company to moderate the
pace of hiring relative to its previous plans.
Evidence strength
CONFIRMED
Why it matters
This is direct company evidence that AI-driven productivity can reduce
incremental demand for human labor without requiring a layoff.
SALESFORCE
2026
Salesforce said AI reduced support demand enough that it no longer needed
to actively backfill support engineer roles.
Mechanism
Non-backfilling / reduced replacement hiring
Evidence
After deploying AI in customer support, Salesforce said support case
volume declined and the company no longer needed to actively backfill
support engineer positions.
Evidence strength
CONFIRMED
Why it matters
This illustrates the non-backfill mechanism directly: positions can
disappear from an organization as employees leave even though no
corresponding AI layoff is announced.
BRINK'S HOME SECURITY
JULY 2026
AI reduced customer-service call volume by roughly two-thirds as the company's
call-center workforce fell from approximately 800 employees to about 400.
Mechanism
Attrition / nonreplacement following AI-driven workload reduction
Evidence
Brink's Home Security reported that AI reduced customer-service call
volume by roughly two-thirds. As the amount of human work declined,
its call-center workforce fell from approximately 800 employees to
about 400. Some employees moved to other positions, while Chief
Information Officer Philip Kolterman said natural attrition accounted
for most of the remaining workforce reduction.
Evidence strength
CONFIRMED
Why it matters
Brink's demonstrates how AI can reduce employment without a corresponding
mass-layoff event. AI reduced the amount of human work required, and
much of the resulting workforce contraction occurred as employees
departed naturally rather than through dismissals.
BANK OF AMERICA
APRIL 2026
Bank of America described attrition and selective hiring as mechanisms for
reducing headcount while using technology and AI to eliminate work.
Mechanism
Attrition / selective replacement hiring
Evidence
CEO Brian Moynihan said Bank of America was reducing headcount through
attrition and careful hiring. He said the company needs roughly 1,300
hires per month simply to keep employment neutral and explained that
management can reduce headcount by controlling hiring and allowing
attrition to reduce the workforce. In the same discussion, Moynihan
described eliminating work through technology and said AI provides
additional opportunities to do so.
Evidence strength
STRONG EVIDENCE
Why it matters
Bank of America illustrates the scale of employment change that can occur
through hiring decisions alone. A large employer can reduce headcount
materially by replacing fewer departing workers, without announcing a
corresponding mass layoff. The evidence connects this strategy to a
broader technology and AI productivity program, although Bank of America
did not attribute every unfilled position specifically to AI.