15MAY

New York moves to count AI job losses

3 min read

15:55UTC

The New York State Assembly passed A 9581 on 3 June, requiring the state Labor Department to report annually on how AI affects hiring, and sent it to the Senate.

← Back to GitLab signs the manifesto, Brussels backs out Jump to analysis ↓

EconomicDeveloping

Key takeaway

New York's Assembly voted to make the state count AI's effect on hiring each year.

The New York State Assembly passed bill A 9581 on Wednesday 3 June 2026, requiring the state Labor Department to publish an annual report on artificial intelligence's effect on hiring and employment ¹. The bill cleared the Assembly and went to the Senate for consideration. It is a measurement mandate rather than a restriction: the state would have to count what AI is doing to jobs, a number no federal instrument currently produces.

The gap A 9581 targets is real. No US federal agency currently measures how many jobs AI removes, which leaves legislators arguing over displacement without an agreed figure. New York's earlier WARN-style disclosure law showed the difficulty: in its first year, 162 companies covering 28,300 workers attributed none of their cuts to AI, because firms decide what reason to state.

The move runs against the opposite direction taken elsewhere. Colorado moved the opposite way in May, replacing its AI Act with a weaker notice-only regime that stripped out the substantive obligations. Two states, two contrary instincts on the same question, with no federal floor to reconcile them. A 9581's value depends on whether the Senate passes it and whether annual reporting can capture a channel that runs mainly through hires never made, the part of displacement that leaves no filing behind.

Deep Analysis

In plain English

The New York State Assembly, the lower house of the state's legislature, passed a bill on 3 June 2026 requiring the state's Department of Labor to publish an annual report on how artificial intelligence is affecting hiring in New York. The bill now moves to the Senate. No US government agency currently produces a regular, official count of how many jobs have been affected by AI. This bill would require New York's Labour Department to start building that record. This bill would require New York's equivalent of the UK's ONS to start keeping that record. Critics point out that a report with no penalties attached and no specific comparison metric may not change employer behaviour. But supporters argue that making the data official and public is the first step toward stronger legislation.

What could happen next?

Precedent
If the New York Senate passes A 9581, it becomes the first US state to create a statutory annual AI-hiring-impact measurement obligation, potentially establishing a template other states adopt.

Source Landscape

This story draws on neutral-leaning sources

Primary parallel: The UK's Disclosure and Transparency Rules introduced in 2006 required listed companies to report executive remuneration in standardised form.

A 2016 review by the High Pay Centre found the reports generated voluminous data but minimal behavioural change until the 2019 gender pay gap reporting requirement added a specific comparison metric and a named publication deadline that created social pressure alongside legal obligation. A 9581's annual report without a comparison metric or mandatory publication deadline follows the pre-2019 UK template rather than the more effective post-2019 one.

Counter-parallel: California's 2002 SB 1 mandatory pay equity reporting requirement was similarly weak on enforcement at passage and was later strengthened by SB 1162 in 2022. The pattern suggests passage of a weak reporting bill is the legislative prerequisite for the stronger enforcement legislation that follows once the data makes the problem visible. A 9581 may be structurally positioned as Phase 1.

Consensus view: The Brennan Center for Justice and Georgetown Law's workplace technology project both describe A 9581's annual-reporting requirement as the minimum viable regulatory response: it generates a factual record without imposing a compliance burden that would trigger industry challenge. The Hawley-Warner federal letter to agencies pursued the same logic at the federal level, and both efforts share the assumption that measurement precedes action.

Counter-view: Cornell ILR School's Alexander Colvin argues mandatory annual reporting without a disclosure trigger date or penalty provision converts the requirement into a box-ticking exercise. New York's prior 2023 AI hiring bias law (Local Law 144) generated extensive reporting but produced almost no enforcement actions, because the reporting requirement did not create a private right of action for affected workers.

The pattern from Local Law 144 is instructive: after nearly a year of operation, zero of 162 mass layoff notices in New York cited AI or automation under the updated WARN Act , indicating firms report what the law technically requires, not what it intends to surface.

Key tension: Whether the annual Labour Department report generates politically actionable data or becomes a statistical annex that legislators cite but cannot act on because the reporting cadence is too slow to match quarterly restructuring cycles.

Sources:Transparency Coalition

Mentions:New York State →California →WARN Act →Brennan Center for Justice →New York A 9581 →

First Reported In

Update #12 · Jobs report says fine, layoff report says no

Transparency Coalition· 8 Jun 2026

Read original →

Causes and effects

Caused by

Colorado guts its AI hiring law

Colorado gutting its AI Act to notice-only sharpens the regulatory divergence that California SB 951's 90-day trigger represents.

Occurred 16 May 2026

Read story →

This Event

New York moves to count AI job losses

It is a bid to build the AI-employment data Congress and the federal statistical agencies have not produced.

Led to

States write the AI law Congress won't

New York Assembly passed A9581 alone on 3 June; Senate passage is the new development that advances the bill to Hochul.

Occurred 10 Jun 2026

Read story →

Different Perspectives

India IT services and global capability centre workforce

India's in-house GCCs added roughly 200,000 net staff in fiscal 2026, nearly double the 110,000 added by the IT services firms feeding the same companies. The shift moves work toward captive centres while squeezing entry-level hiring at the outsourcing firms, reshaping where Indian tech careers begin as US clients cut staff at home.

EU workers and European labour institutions

The 93-4 committee vote locked the diluted Omnibus literacy clause before plenary: EU workers in AI-augmented but non-high-risk workplaces have no statutory right to demand an explanation until December 2027. The European Trade Union Confederation called the shift from 'ensure' to 'support' a legal threshold collapse, not a drafting compromise.

UK workforce and labour market

UK 16-to-24 unemployment reached 16.2% in the latest ONS reading, above the 15.2% pandemic peak and the highest since 2015. Britain is among the most AI-exposed labour markets this desk tracks, yet the Office for National Statistics still publishes no AI-attribution layer, so young workers face the displacement without official data naming its cause.

Anthropic and frontier AI labs subject to US jurisdiction

Anthropic complied with the directive but publicly disputed its application, citing that OpenAI's GPT-5.5 carried the identical jailbreak vulnerability and remained on sale. For any US-domiciled frontier lab, the action demonstrates that regulatory compliance and political alignment are now distinct variables: Anthropic backed the pro-regulation PAC and was the first lab Washington reached.

US national-security and export-control apparatus

The Lutnick directive treats runtime inference access by a foreign national as legally equivalent to exporting Claude Fable 5 and Mythos 5 to that person's home country. It established that a deployed consumer AI product can be withdrawn globally by regulatory letter, with no appeal period and no customer notice.

European workers and regulators

NBER working paper w34995 found European workers use generative AI at 32% versus 43% of US workers, a gap driven by management practice rather than regulation. The EU AI Act's high-risk employment deadline stays at December 2027, leaving European workers facing the same displacement curve two to four years behind the US.