Layoff notices, SEC filings, bankruptcies, jobless claims and JOLTS. One search. One API.
LaborCurrent collects WARN layoff notices from 47 states and DC, SEC 8-K filings, bankruptcy petitions, DOL jobless claims and JOLTS turnover — plus H-1B/LCA petitions as the hiring counter-signal — into one normalized dataset, each record carrying the agency and dataset it came from. Search it free, plug it into your product through the API, or buy the data outright.
- Product & data teams REST API, MCP server, webhooks, exports
- Investors & analysts employer distress signals, watchlists, Risk Signal API
- Researchers, journalists & lawyers citable records, company histories, one-time datasets
See how to research a layoff filing
Follow a market down to a single company record, then open the public source the notice came from. No account needed to watch.
Walkthrough · 2:40 · recorded 24 September 2026
Explore the dataRead the transcript
You have a company or a region to understand. What has actually been filed? Where are the jobs affected? And which source should you read next? LaborCurrent brings public workforce records into a place you can explore. Let's start with one practical task.
Finding layoff notices in a market you care about. Then following the evidence behind them. Here is the layoff explorer. I'll choose Texas and the table moves from states to counties. Now we can narrow the question instead of treating a national total as our answer. The columns show notices, affected workers, and the dates represented.
Those dates matter. An archive can help with history while a recent filing answers a different question. Keep your research period in view. From here, work toward the company or location you need. A notice gives you a starting point. The employer, the place, the reported headcount, and the dates.
Read notice dates and effective dates separately whenever they are available. A filing date is not automatically the day jobs end. And several rows under one company do not automatically represent separate groups of workers. Check the underlying records before adding them together. The next step is the source. Follow the agency link and check the record in its original context.
This is where you confirm the employer name, location, and dates before quoting a number or sharing a finding. LaborCurrent helps organize the search. Your research trail should still include the public source. Save the page and the date you checked it alongside your notes. A WARN notice is only one kind of record. The site also provides SEC filings, bankruptcy records, H-1B, and LCA petitions, unemployment claims, and JOLTS data.
Each answers a different question. Company filings can add employer context. Claims and JOLTS describe broader labor conditions. A petition is not proof of a completed hire. And records appearing near each other do not establish what caused a layoff. Before taking the analysis further, review coverage and methodology.
The available history, freshness, and delivery options depend on the product you use. You can browse, investigate API access, or look at a dataset purchase for offline work. Start with a company or a location you already understand. Find a record, check its source, and see whether this makes your next research task easier. That is the best first test of LaborCurrent.
One employer, connected
What you see before you pay for anything. Oracle, as the agencies published it, as of 2026-09-19: the notices, the filings and the petitions, each attributed to the agency that published it.
Oakland, CA · 279 + 99 workers, same notice date. 23 notices on record since 2024-02, 3,518 workers. Sources: WA ESD, CA EDD.
The same week as the notices. Source: EDGAR, 2026-09-14 · EDGAR, 2026-09-10.
The hiring counter-signal beside the cuts: roles, wages and worksites by city. Source: DOL LCA disclosure files, imported weekly.
Open Oracle's company page → · Every company → · How names are matched →
6 datasets. One place.
Workforce change shows up in a sequence. Each dataset catches a different moment; every night we collect what the agencies publish, normalize it and join it on employer, place and date — and the joined record set is what becomes intelligence.
Open a dataset from the diagram, or search everything → · Coverage, cadence and known gaps →
What you can do with it
Build a product
One REST API over six datasets, JSON responses, an MCP server for AI agents, and CSV/JSON/Parquet exports on paid plans. Deterministic record IDs, so your integration can re-fetch safely.
Research companies
Search WARN notices by company, city or state; open a company page to see every filing under its name variants next to its SEC 8-K, bankruptcy and H-1B/LCA records. Buy a full dataset when you need the whole history offline.
Monitor employers
Free email alerts for WARN notices and SEC 8-K filings, with a three-company watchlist on a free account. Pro adds webhooks; Investor adds the Risk Signal API and a 100-company watchlist.
The latest filings, and your first API call
Latest filings
The newest notices on record, exactly as the public API returns them to a visitor without a key: the last 30 days, published with the free-data delay. Every row links to the company and location pages.
These rows are published with the free 7-day delay. Same-day WARN notices are on Starter, $49/month →
Inspect sample rows → · All recent layoffs → · Browse by state → · Free email alerts →
Your first request
Every call goes to https://laborcurrent.com/api/ with your key in the X-API-Key header. Aggregate endpoints such as /api/stats need no key.
# Fetch the latest 10 layoff notices
curl -H "X-API-Key: YOUR_KEY" \
"https://laborcurrent.com/api/records?limit=10"
# California notices from 2026
curl -H "X-API-Key: YOUR_KEY" \
"https://laborcurrent.com/api/records?state=CA&date_from=2026-01-01"
JSON responses, deterministic record IDs, filters by state, company, date and headcount; an MCP server so AI agents can use the same data.
Need the whole table, not calls? CSV and JSON exports are on Starter; Parquet on Pro; or buy a one-time dataset →
Full quickstart → · Endpoint reference → · MCP for AI agents → · How records are normalized →
Coverage and freshness
Daily WARN notices
Scraped daily at 05:00 UTC from 47 state workforce agencies and DC. Arkansas, Wyoming and New Hampshire do not publish WARN data; we say so rather than fill the gap. Company names are normalized; the raw filed name and source URL are kept on every record.
Daily · weekly · monthly Connected filings
SEC 8-K Items 1.03 and 2.05 and federal bankruptcy petitions are refreshed daily; DOL unemployment claims weekly; H-1B/LCA weekly for the current fiscal year with a monthly full import; JOLTS monthly, as the agencies publish.
Evaluation tier Trial limits
Without a key: the last 30 days, published with a 7-day delay. An evaluation key: 10 calls/day, WARN only, 30-day delayed data. Upgrade to Free (no card) for 25 calls/day and all six datasets. Starter makes WARN live; Pro removes all delays — see pricing.
AI and the workforce
AI is taking jobs off the books. The record shows where, and who is hiring instead.
Automation and AI decide which jobs are posted, moved or cut. We do not guess at causes: every WARN notice, 8-K and visa petition is shown as the agency published it, with its source and date — so when an employer closes a site and files for AI talent the same quarter, you see both, by name.
WARN notices name the site, the headcount and the date; 8-K Item 2.05 filings put the restructuring charge on the record.
H-1B and LCA petitions show who is hiring machine-learning and data roles, at what wage, in which city; JOLTS shows the sectors still opening jobs.
Investors position ahead of a restructuring and separate the employers cutting from those investing; recruiters see released skills by metro; workforce teams plan; journalists and lawyers cite the filing itself.
Risk Signal API → · For investors → · Research and analysis → · Daily market brief → · Affected by a layoff? Get help → · Partner inquiry →
Built for the people who act on it
Immigration policy is an investment research question. Explore employer-level sponsorship and workforce records — H-1B petitions, LCA filings and WARN notices, side by side.
For investors →Pick a plan when you have seen the data
- 25 calls/day
- All six datasets, 7-day delay
- 90 days of history
- Email alerts · 3-company watchlist
- 200 calls/day
- Live WARN, other datasets 7-day delayed
- CSV / JSON export
- 1 year of history
Compare plans → · One-time datasets from $49 →
Start with an evaluation API key
10 calls/day, WARN dataset only, 30-day delayed data. Upgrade to Free (no card) from your account for 25 calls/day, all datasets, and dashboard access — see pricing.
10 calls/day · WARN data · No credit card · Verify by email to activate