Compsci grads facing recession-like job prospects thanks to AI

US Census Bureau economists found that graduates in the top decile of AI-exposed majors saw initial employment likelihood fall 5 percentage points and full-quarter initial earnings decline 13% following ChatGPT's late-2022 release. About half of the earnings decline reflects movement into lower-paying industries such as retail and food service, with computer science, mathematics, statistics, some engineering fields and journalism among the most affected. Less AI-exposed fields including nursing and education did not show comparable early-career deterioration, though researchers caution against extrapolating the short-term results into a broader labor-market forecast.

Analysis

The investable implication is not a broad consumer-demand shock but a gradual transfer of value from entry-level knowledge labor to software vendors and employers with scalable workflows. Staffing and outsourced IT vendors—ASGN, RHI, KFY, EPAM, and GLOB—face the clearest 6-18 month risk: lower junior requisition volumes reduce utilization and pricing, while clients can use AI as leverage in rate negotiations. Conversely, Microsoft, ServiceNow, and Salesforce can convert labor-substitution narratives into seat expansion and retention, but the earnings benefit will be delayed because their own R&D and sales organizations remain labor-intensive.

The key contrarian point is causality: early-career technical hiring is also highly sensitive to venture funding, interest rates, and the post-2021 normalization in tech headcount. If aggregate tech job openings recover without a corresponding recovery in junior roles, that would validate structural substitution and justify multiple compression for staffing; a broad rebound in requisitions would instead identify this as a cyclical trough. Lower-wage retailers and restaurants may receive a modest labor-supply benefit, but this is unlikely to move earnings for WMT, MCD, or YUM absent evidence of wage-rate moderation.

Near term, this is primarily a sentiment and labor-data watch item rather than a catalyst for mega-cap AI longs already priced for large productivity gains. The more actionable second-order risk is that firms monetizing human-hours, rather than software licenses, may see estimates fall before reported revenue does as management teams reset hiring and utilization assumptions.

Key Decisions for Investors

• Establish a 6-12 month pair: long MSFT / short ASGN, sized modestly. The thesis is relative operating leverage to AI-enabled productivity versus exposure to lower technical-placement demand; exit if US tech job openings and entry-level requisitions recover together for two consecutive monthly readings.
• Maintain an underweight or put-spread watch on RHI and KFY into the next two earnings cycles. Require confirmation through declining placement volumes, weaker consultant utilization, or guidance cuts before adding; a sharp rate-cut-driven white-collar hiring rebound is the principal risk.
• Do not chase broad AI software on this labor datapoint alone. Add to NOW or CRM only if upcoming results show measurable customer workflow consolidation, improving subscription margins, or reduced services intensity rather than generic AI-product adoption claims.
• Monitor NFP wage growth for 20-24 year-old workers, Indeed/LinkedIn software-engineer postings, and university enrollment trends over the next 1-3 months. A divergence of stable senior hiring with continued junior-job weakness is the signal to increase the staffing short exposure.