{"slug": "the-state-of-ai-agents-in-numbers", "title": "The state of AI agents, in numbers", "summary": "Nearly two-thirds of all tracked AI job postings specifically build or run agents, according to Get Ready For Agents' live indexes covering 216 companies actively hiring for agent roles from a tracked field of 261. Of 3,164 tracked postings, only 40 are for agent governance or security, highlighting a lag in controls behind building. Across the industry, 62% of organizations are experimenting with AI agents but only 23% are scaling them, and over 40% of agentic AI projects are expected to be scrapped by 2027, per Gartner.", "body_md": "**Quote us in one click.** Every figure below is dated and sourced. Hit Copy quote on any stat for a ready-to-paste line with its attribution built in. Our index figures are published under CC BY 4.0.\n\nOur own research\n\n## The numbers only we have\n\nComputed live from the Get Ready For Agents indexes — updated automatically, so a quote from this page is never stale. These are the figures that put your name in the citation.\n\nAcross 216 companies actively hiring for them, from a tracked field of 261.\n\nNearly two-thirds of all tracked AI job postings specifically build or run agents.\n\nAt AI-native companies, agent work dominates the engineering hiring mix.\n\nOf 3,164 tracked postings, only 40 are for agent governance or security — the controls lag the building.\n\nAgent teams are overwhelmingly being built in the office, not remotely.\n\nLive across 16 countries and 4 regions in the Governance Index.\n\nPlaced across four quadrants of the agent-building landscape in the Agent Index.\n\n### More ways to break it down\n\nThe same live index, sliced four ways. Every breakdown copies as a single attributed line.\n\n#### AI hiring by discipline\n\nOf 3,164 tracked AI postings — building roles dominate; the control functions barely register.\n\n#### Who is hiring the most\n\nCompanies ranked by open agentic-role count.\n\n#### The skills agents demand\n\nMost-requested skills across agentic roles.\n\n#### Frontier labs vs. enterprise\n\nShare of each cohort's AI roles that are specifically agentic.\n\n### Australia vs. the USA\n\nThe same index, split by country. Australia's AI hiring skews *more* agentic than the United States.\n\n**55** tracked AI roles\n\n**69%** are agentic\n\n**28** companies hiring\n\n**1,481** tracked AI roles\n\n**63%** are agentic\n\n**154** companies hiring\n\nThe wider picture\n\n## The industry, from high-authority sources\n\nA deliberately small, curated set — named analysts, research bodies and operator-reported production numbers only. Thin or unverifiable claims are left out on purpose, so everything here is safe to cite.\n\nAdoption & the reality gap\n\n### Everyone is trying agents. Far fewer are scaling them.\n\nThe headline from every major 2025–26 survey is the same: experimentation is near-universal, but production and measurable value are still the exception.\n\nOver 40% of agentic AI projects will be scrapped by the end of 2027, driven by escalating costs, unclear business value, and inadequate risk controls.\n\n62% of organisations are at least experimenting with AI agents, but only 23% are actively scaling them.\n\nJust 39% of organisations deploying AI agents report a measurable impact on EBIT.\n\nGovernance & oversight\n\n### The controls are lagging the deployments.\n\nAgents are reaching production faster than the governance, evaluation, and oversight around them — the exact gap our own indexes were built to measure.\n\nOnly one in five organisations has a mature governance model for autonomous AI agents.\n\n89% of teams building agents have implemented observability, but only 52% have implemented evaluations — a 37-point gap between watching agents and testing them.\n\nROI in production\n\n### Where agents are working, the numbers are large.\n\nOperator-reported results from named enterprises running agents at scale. Each figure is self-reported by the company named — quote them as such.\n\nKlarna's AI assistant handled 2.3 million customer conversations in its first month, cutting average resolution time from 11 minutes to under 2.\n\nJPMorgan's COiN platform reviews 12,000 commercial credit agreements in seconds — work that previously consumed roughly 360,000 lawyer-hours a year.\n\nSalesforce's Agentforce resolves around 85% of incoming customer service requests autonomously on Salesforce's own help portal.\n\nMorgan Stanley's DevGen.AI reviewed 9 million lines of legacy code, saving an estimated 280,000 developer-hours.\n\nOur own figures (Jobs, Governance and Agent indexes) are primary data published by Get Ready For Agents under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). External figures are attributed to their original source — cite that source directly. Operator-reported results are self-reported by the company named.\n\n## Agent statistics, answered.\n\nThe figures people ask us for most — with the current numbers baked in.\n\n## How many open AI agent jobs are there?\n\nAs of July 2026, the Get Ready For Agents Jobs Index tracks 2,071 open agentic AI roles across 216 companies — about 65% of all tracked AI hiring. The figures update automatically from live job postings.\n\n## What share of AI hiring is specifically for agents?\n\nAround 65% of tracked AI job postings are specifically agentic — building or running AI agents rather than general machine-learning work. At frontier AI labs that rises to 68%.\n\n## How many AI governance frameworks are there?\n\nThe Get Ready For Agents Governance Index tracks 35 AI governance frameworks in force across 16 countries and 4 regions, updated monthly from official regulator and standards-body sources.\n\n## Can I cite these statistics?\n\nYes. Every figure on this page is dated and sourced. Our own index figures (Jobs, Governance, Agent) are published by Get Ready For Agents under CC BY 4.0 — quote them with attribution. External figures carry their original source; cite that source directly.", "url": "https://wpnews.pro/news/the-state-of-ai-agents-in-numbers", "canonical_source": "https://www.getreadyforagents.com/statistics/", "published_at": "2026-07-25 22:06:28+00:00", "updated_at": "2026-07-25 22:22:20.202540+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents"], "entities": ["Get Ready For Agents", "Gartner", "Klarna", "JPMorgan", "Salesforce", "Morgan Stanley"], "alternates": {"html": "https://wpnews.pro/news/the-state-of-ai-agents-in-numbers", "markdown": "https://wpnews.pro/news/the-state-of-ai-agents-in-numbers.md", "text": "https://wpnews.pro/news/the-state-of-ai-agents-in-numbers.txt", "jsonld": "https://wpnews.pro/news/the-state-of-ai-agents-in-numbers.jsonld"}}