Part of our AI Regulation and Compliance News series
Connecticut Just Made You Put AI on the Layoff Paperwork
On May 27, 2026, Governor Ned Lamont signed Connecticut SB 5, the state's Artificial Intelligence Responsibility and Transparency Act. Most coverage framed it as one more state AI law in a crowded year. For anyone who runs a reduction in force, that framing misses the part that actually changes your job. Beginning October 1, 2026, when a covered Connecticut employer files a layoff notice under the state WARN Act, it has to state whether the workforce reduction is connected to the employer's use of artificial intelligence or other technological change.
Read that again. The robot question is now a box on the layoff filing. You can no longer quietly call a reduction "restructuring" if automation is what drove it. The filing forces the answer, and once it is filed, it is frozen, dated, and discoverable. That is a different world from the one most HR teams have been operating in.
Key takeaways
- Starting October 1, 2026, Connecticut WARN Act layoff notices must disclose whether the reduction relates to the employer's use of AI or other technological change.
- It binds employers with 100 or more employees, the same threshold that already triggers WARN notice duties in the state.
- The Connecticut Attorney General enforces it as an unfair or deceptive trade practice, so a sloppy or evasive answer is its own exposure.
- A second wave on October 1, 2027 adds pre-decision notice for automated employment decision tools and removes AI as a defense to discrimination claims.
What SB 5 actually requires
The headline change for HR is narrow and concrete. Connecticut's existing WARN Act already requires advance notice of large layoffs and plant closings. SB 5 adds a new field to that notice. When you file, you have to say whether the layoff is related to the employer's use of AI or other technological change. It is a disclosure duty, not a ban. Connecticut is not telling you that you cannot automate a function and reduce the headcount that did it. It is telling you that if that is what happened, you have to write it down on the record.
That sounds small until you sit with the consequences. The stated reason on a layoff filing has always mattered in litigation and agency review. Now there is a specific, dated, government-filed admission about whether technology caused the cut. If an affected employee later claims the reduction was a cover for age or disability discrimination, your WARN disclosure is one of the first documents that gets pulled. The answer you give in a hurry on October 2 is the answer you defend two years later.
Who and what it covers
The threshold tracks the state WARN Act: employers with 100 or more employees, filing notice for the qualifying mass layoffs and closings that already trigger WARN obligations. If you are big enough to owe a Connecticut WARN notice today, you are inside this rule on October 1. The disclosure attaches to the layoff event, so the trigger is the reduction itself, not whether you think of yourself as an "AI company." A logistics firm that swaps a scheduling team for software is as covered as a tech employer.
"Other technological change" is deliberately broader than AI alone. A reduction driven by new automation, a workflow platform, or a robotics deployment can fall inside the disclosure even if no one on your team would call the tool artificial intelligence. The practical test for HR is causation, not branding: did a technology decision drive this reduction, in whole or in substantial part. That is the question you have to be able to answer honestly and consistently before the notice goes out.
The October 1, 2026 deadline and who enforces it
The deadline is not soft. October 1, 2026 is the earliest hard compliance date of any state AI-and-employment law on the books, which is why it deserves a place at the top of your calendar rather than your watch list. The Connecticut Attorney General enforces SB 5, and the enforcement frame is the state's unfair and deceptive trade practices law. That matters. An evasive or false disclosure is not a paperwork foot fault; it is the kind of conduct that statute is built to punish, with the AG, not a private plaintiff, deciding whether to act.
So you have two distinct risks on the same filing. Get the disclosure wrong by understating AI's role and you have a potential deceptive-practices problem with the AG. Get the underlying business reason wrong and you have handed a discrimination plaintiff a frozen, dated document. The safe path runs between them: an honest, accurate, defensible answer arrived at before the filing, not improvised at the deadline.
Connecticut is not standing alone here, which is the part that should pull this off your back burner. Colorado's AI employment law and the federal direction described in the EEOC's enforcement plan for AI hiring tools point the same way, and the broader duty to give workers honest notice when AI shapes an employment decision is the subject of our briefing on AI layoff notice law. Connecticut just put the soonest hard date on it.
A compliance checklist to run now
You have months, not years, and the answer you file is hard to unwind. Build the process before October, not on the day a reduction lands.
- Map every Connecticut headcount decision in your pipeline against the 100-employee WARN threshold so you know which events are in scope before they happen.
- Run a written causation screen on each planned reduction that asks whether AI or other technology was a substantial factor, and keep the analysis, not just the conclusion.
- Decide the disclosure answer at the planning stage, with legal in the room, so the WARN filing reflects a considered position rather than a deadline guess.
- Pressure-test the stated business reason against the real record before you file, because once it is filed it freezes and becomes evidence.
- Document the human decision-maker on every reduction so you can show a person owned the call, not a model output.
- Brief your Connecticut managers now, because the people who design the reduction are the ones who know whether technology drove it.
What HR owes now
The job has shifted from explaining a layoff after the fact to documenting its cause before the fact. That is a real change in the order of operations. You owe the company a process that produces an accurate disclosure under deadline pressure, and you owe affected employees an honest answer about what drove the decision. Both duties point to the same discipline: figure out, in writing and early, whether technology caused the cut, and say so plainly.
This is exactly the kind of repeatable, defensible workflow you can build with an AI assistant working for you instead of against you. A short pre-reduction screen run with Claude can help you frame the substantial-factor question, draft disclosure language that is candid without being self-incriminating, and surface gaps in the stated business reason while you can still fix them. The tool does not decide. It makes the human decision faster, more consistent, and better documented. That is the entire premise of The Leveraged HR Professional, our course on building AI-assisted HR systems that hold up under exactly this kind of scrutiny.
What AI does not replace
Be precise about the line. AI can help you draft the disclosure and stress-test the reasoning. It cannot decide whether to run the reduction, it cannot be the decision-maker of record, and it certainly cannot sign the WARN filing for you. SB 5 and the laws moving alongside it are tightening, not loosening, the requirement that a human owns the employment decision. A disclosure generated by a tool and filed without a person standing behind it is the worst of both worlds: automated and unaccountable.
So use the assistant to do the drafting, the screening, and the consistency checks. Keep the judgment, the accountability, and the signature with a named human being. The point of getting good with these tools is not to hand the decision to software. It is to walk into October 1 with a process clean enough that you would be comfortable explaining it to the Attorney General.
Not sure which AI skills your role needs first?
Take the two-minute course finder and get a path built for your work.
Join The Leverage Club for $49.
Working sessions, templates, and briefings like this one for senior professionals putting AI to work the right way.
Frequently asked questions
Does Connecticut SB 5 ban using AI to reduce headcount?
No. SB 5 does not prohibit automating a function or reducing the team that performed it. The October 1, 2026 change is a disclosure duty. When you file a Connecticut WARN Act notice, you have to state whether the reduction is related to the employer's use of AI or other technological change. The law regulates what you say on the record, not whether you can make the decision.
When does the AI layoff disclosure requirement take effect, and who has to comply?
The disclosure requirement begins October 1, 2026. It applies to employers covered by Connecticut's WARN Act, generally those with 100 or more employees, when they file notice for a qualifying mass layoff or closing. A second wave on October 1, 2027 adds pre-decision notice for automated employment decision tools and removes AI as a defense to discrimination claims.
What happens if the disclosure is wrong or evasive?
The Connecticut Attorney General enforces SB 5 as an unfair or deceptive trade practice, so an inaccurate or misleading disclosure can be its own violation. Separately, because the WARN filing is dated and discoverable, an inaccurate business reason can surface later in a discrimination claim. The safer approach is to decide the answer before filing, document the causation analysis, and keep a named human decision-maker on the record.
Related briefings: Colorado AI hiring law: what HR owes now ยท The EEOC enforcement plan for AI hiring tools ยท AI layoffs and employee notice law