The AI boom hasn’t stopped U.S. companies from hiring cheap offshore labor, and overseas call center employment is still skyrocketing

Salesforce cut 4,000 customer service roles in Sept 2025, but new cross-country labor data suggests offshore call-center work has not been displaced by AI. Apollo chief economist Torsten Slok cites Philippines call center employment nearly doubling to ~2 million from 2016-2025 and unemployment falling in the Philippines (9% to ~5%) and India (~7% to ~6%) through July 2026, implying AI is expanding rather than eliminating demand. The article also notes an AI conversational tool boosted customer support agent productivity by ~14% per hour (2023 Stanford research), consistent with a “Jevons paradox” dynamic where lower interaction costs increase total volume served.

Analysis

The important signal is not displacement speed but demand elasticity: if AI lowers the cost of handling one interaction, management usually discovers more interactions worth selling, serving, or outsourcing. That is bullish for platforms that monetize workflow volume rather than headcount reduction, and it argues against treating layoff announcements as evidence of durable margin savings. For CRM, the key question is whether agentic features expand paid usage and consumption enough to offset pricing pressure; if they do not, the AI story stays a narrative premium rather than an earnings driver.

Second-order, the structural beneficiaries may be the labor-arbitrage operators and cross-border commerce rails rather than the domestic workers the market keeps focusing on. Offshore BPO capacity can become a demand sponge when AI improves throughput but still needs human escalation, which should support utilization before it meaningfully compresses wages. EBAY is a mild beneficiary through lower-friction translation, support, and seller onboarding, but the revenue impact is more likely to show up as incremental conversion and retention over 1-3 quarters, not a step-change in multiple.

The contrarian miss is that consensus is too focused on task automation and not enough on channel expansion and service intensification. The thesis is falsified if we see actual headcount contraction at major offshore BPOs, falling utilization, or CRM-style AI attachment failing to lift consumption metrics over the next two earnings cycles. Until then, the burden of proof is on the bears: AI may be reducing unit cost while expanding total labor demand, not shrinking it.

Key Decisions for Investors

• Buy CRM on pullbacks only if next-quarter commentary shows higher AI-assisted usage/consumption; 3-6 month upside is intact, but cut quickly if AI modules remain seat-based and fail to move net retention or cRPO.
• Long EBAY as a modest 1-2 quarter beneficiary of AI-enabled translation/support and seller conversion; risk/reward is better as a small tactical add than a core thematic bet because the impact is likely incremental, not transformative.
• Do not short offshore services on the layoff narrative alone; instead, place on watchlist any public BPO names or proxies if they print rising utilization with flat-to-up headcount, which would confirm the Jevons dynamic.
• No direct trade in TSCC or VABK from this note; the read-through is too weak and any position would be thesis drift.
• If CRM AI monetization fails to accelerate by the next two earnings prints, fade the multiple premium and reduce exposure; that would indicate the market is paying for cost takeout that never shows up in revenue.