Rahul SomvanshiAmerican companies cut more than 117,000 jobs this year citing artificial intelligence, eliminating an estimated $8.5 billion in annual wages, in a wave of cuts that researchers are now tracking against actual AI performance data. That data tells a different story than the executive announcements.
The Remote Labor Index, which measures how often AI agents complete real freelance projects at professional quality, found that Claude Fable 5 — the leading model on the index — completed 15.8% of assigned projects. GPT-5.5 reached 6.3%. The index, developed jointly by Scale AI and the Center for AI Safety, tests AI agents on real freelance work categories including web development and video production. The gap between what executives cut for and what AI can currently deliver is the central finding in reporting from The Guardian and the AI Wage Loss Dashboard tracked through October 2026.
If your job was eliminated as part of an AI-automation announcement, or if your company has described AI as a replacement for your role, the performance figure matters. A tool that completes fewer than two in ten professional freelance projects independently is not a full substitute for a skilled worker. A Guardian investigation found that 90% of surveyed executives said their AI initiatives had not produced measurable productivity gains.
The Performance Gap Driving the Backlash
The 15.8% completion rate applies to freelance project tasks, not all workplace activities, and the index methodology acknowledges the figure reflects current capability across the tested task set. Not all tasks are equal in complexity, and performance will vary by category. What it does establish is that the automation case many companies used to justify layoffs assumed a capability level that does not yet consistently exist in commercial tools.
The legal system has started pricing that gap. During a Harris County trial in August, a 3M expert witness named Josh Autenrieth was found to have used ChatGPT to draft a defence report — instructing the tool to show that 3M bore no fault for a deadly industrial explosion. Plaintiffs’ lawyers obtained roughly 350 pages of his chat logs in discovery, showing prompts asking ChatGPT to build a zero-fault case. A jury subsequently awarded $61.5 million to 24 plaintiffs, finding 3M 30% responsible for the 2020 Watson Grinding explosion in Houston. 3M said it disagrees with the verdict and plans to appeal. AI in legal proceedings — what the courts have decided
For workers in affected sectors, the performance gap creates a specific job-market dynamic. Roles were cut on the assumption AI would replace them; AI is not reliably replacing them at the levels projected; but the headcount is gone.
Which Jobs Are Being Tracked and Where
The AI Wage Loss Dashboard, published by the Center for AI Safety at wageloss.ai, tracks announced job cuts attributed to AI and estimates the associated wage impact nationally. The dashboard is a reporting tool, not a causal analysis — it aggregates employer-disclosed and reported cuts attributed to AI, but does not independently verify that AI has replaced those roles or that the tools are performing as claimed.
The $8.5 billion wage-loss figure represents the estimated annual wages eliminated through the tracked period across the United States. how to read AI layoff tracker data
The Remote Labor Index performance data is particularly useful context for workers deciding how to position their skills. If AI tools are completing 15.8% of real-world freelance projects, the remaining tasks still require human judgement, verification, or execution — and employers who cut aggressively are finding that out through productivity reports, not press releases.
Frequently Asked Questions
Does an AI layoff tracker prove AI caused my layoff?
No. Trackers like the Center for AI Safety’s wageloss.ai dashboard and other national tools record job cuts that employers attributed to AI automation. They are early-indicator datasets, not causal analyses. A layoff recorded in the tracker means the employer cited AI; it does not verify that AI has actually replaced the role or that the replacement is performing as claimed.
If AI tools only complete 15.8% of freelance projects, why are companies still cutting jobs?
The cuts largely preceded the performance data. Companies announced automation timelines based on projected AI capability, not measured performance on real work. The high rate of executives reporting no productivity gains suggests many companies are now operating with reduced headcount and tools that are not covering the gap.
What should workers in AI-exposed roles do with this data?
Track both the capability data (Remote Labor Index, independent benchmarks) and your company’s actual deployment. If AI tools are handling a specific, measurable portion of your workload, that is useful to know. If the tools are announced but not performing, the productivity gap is an argument for your continued role — and one you can make with data rather than opinion.
\———————————————————————————————–
\———————————————————————————————–
## How to Log Into WordPress REST API — Step by Step
### What We Used: \*\*Application Password Authentication\*\*
\—
### Step 1: Get the Credentials
From the Pipeline file (\`Karmactive Pipeline.md\`), the credentials are:
\- \*\*Username\*\*: \`Sunita Somvanshi\`
\- \*\*Application Password\*\*: \`Tw8Q Khc9 8KxU 5L4O 3Q0s 0w6J\` (remove spaces → \`Tw8QKhc98KxU5L4O3Q0s0w6J\`)
> Application Passwords are created in WordPress Admin → \*\*Users → Profile → Application Passwords\*\* section. They’re separate from the login password and designed for API access.
