What 180M Job Postings Reveal About AI’s Real Impact on Jobs in 2025

Analyzing 180M postings, the author finds an 8% overall decline in 2025, with AI’s impact concentrated rather than universal. Creative execution roles fall sharply while strategic creative, leadership, software engineering (overall), and customer service remain comparatively resilient; AI-focused engineering and infrastructure surge. Regulatory shifts, not AI, drove steep drops in compliance/sustainability, and influencer marketing stands out as a growing marketing niche.

Key Points

Overall job postings fell 8% YoY in 2025, but declines and gains vary widely by role, revealing selective, not universal, AI impact.
Creative execution roles (e.g., CG artists, writers, photographers) are in two-year decline, while creative leadership and decision-heavy design roles are more resilient.
Regulatory pullbacks drove steep drops in compliance and sustainability jobs across seniority, while trade compliance rose amid tariff activity.
AI-talent and infrastructure are surging: ML engineers (+40%) lead growth, alongside robotics, research/applied scientists, and data center engineers; leadership roles outperform managers and ICs.
Software engineering demand is broadly steady (frontend softer), and customer service reps are not being mass-replaced by AI; influencer marketing is a bright spot within marketing.

Sentiment

The HN community was predominantly skeptical of the article's methodology and causal claims, while finding some of its directional observations plausible. Commenters generally agreed that AI is affecting certain job categories but pushed back hard on the analysis's ability to isolate AI as the cause versus macroeconomic factors, post-pandemic corrections, and political or regulatory shifts. The discussion was substantive and data-literate, with many commenters offering alternative explanations and demanding more rigorous analysis. The article's author earned goodwill by engaging constructively and updating the post in real time.

In Agreement

Frontend engineering is more susceptible to AI automation than backend because UI code is repetitive, mistakes have small blast radius, and JavaScript has abundant training data for LLMs.
Software engineering demand is buoyed by induced demand — as AI makes software cheaper to produce, organizations want more and more complex software, offsetting productivity gains.
Creative execution roles are genuinely declining while creative leadership and strategy roles hold up, reflecting AI's ability to handle rote production tasks but not high-level creative direction.
AI coding tools currently boost engineer productivity rather than replacing engineers, consistent with the article's finding that software engineering postings held near baseline.
Senior leadership roles are more resilient because they require judgment, strategy, and human decision-making that AI cannot replicate.
ML engineer demand is genuinely surging, driven by massive corporate investment in AI infrastructure.

Opposed

The analysis lacks absolute job counts, making percentage changes misleading — a large percentage gain in a tiny job category may be far less impactful than a small percentage decline in a massive one.
Two years of data is insufficient to attribute trends to AI specifically, given confounding factors like post-pandemic normalization, recession, tariffs, and the end of zero-interest-rate hiring.
The article conflates correlation with causation — job posting declines in areas like nursing, sustainability, and compliance have obvious non-AI explanations such as political backlash and regulatory changes.
Job postings are a poor proxy for actual employment; the data is heavily polluted by ghost jobs, duplicate listings, and aggregator effects, and the claimed job count may be significantly inflated.
The broader economy may be in recession when the AI bubble is removed, making broad job declines an economic phenomenon rather than an AI-driven one.
Big Tech is offshoring jobs to India at scale, and US job posting declines may reflect geographic shifts rather than AI replacement.
Companies are using AI is replacing jobs as a convenient narrative to justify layoffs driven by cost-cutting and post-pandemic over-hiring corrections.