When AI drives the layoff, you may now owe the workers notice.
Connecticut just made notice mandatory when AI causes a reduction in force. Here is the calm, practical read for people-ops, including why this is a template other states will copy.
Key Takeaways
- What changed: in late May 2026, Connecticut Governor Lamont signed an omnibus AI law that restricts employer AI use and, uniquely, requires employee notice when AI causes a reduction in force. The signing is confirmed by the Governor's office, and the employment provisions were flagged by Ogletree.
- The new category: most layoff law is about headcount thresholds and timing. This is different. It attaches a duty to the cause of the cut. If AI made or shaped the decision, the worker may be owed formal notice about it.
- Why it touches you even outside Connecticut: AI is now the leading cited reason for layoffs in 2026, per Forbes reporting. A duty that follows the cause, not the headcount, is exactly the kind of rule that spreads, so the smart move is to build the habit before your state copies it.
- The real takeaway: this is not a reason to stop using AI in workforce planning. It is a reason to document how you used it, so that when notice is required you can give an honest, defensible account of the decision.
The Leveraged Years Briefing. Permalink
What Connecticut actually passed
In late May 2026, Connecticut Governor Ned Lamont signed an omnibus artificial intelligence law. The signing is on the record from the Governor's office, and the employment side of the bill was broken down by the law firm Ogletree. The headline for people-ops is one specific provision: when an employer's use of AI causes a reduction in force, the affected employees are owed notice.
That sounds small. It is not. Almost every layoff obligation HR already knows is keyed to numbers and calendars. How many people, in what window, with how many days of warning. Connecticut added a different kind of trigger. This one is keyed to why the cut happened. If AI drove or materially shaped the decision to let people go, a notice duty attaches to that fact.
The rest of the law restricts employer AI use more broadly, which is the part that lines up with what other states have been doing. The notice-on-AI-layoff piece is the part that is genuinely new, and it is the part worth getting ahead of.
Why this is not just a Connecticut problem
Here is the context that makes a single state law worth your attention. Forbes reporting this year shows AI is now the most commonly cited reason for layoffs in 2026. The thing Connecticut just regulated is not a fringe scenario. It is the reason a growing share of cuts are happening.
When a regulator writes a rule that targets the most common version of an event, that rule tends to travel. Connecticut has given the other states a finished template: a clean, copyable concept that ties a worker protection to AI as the cause of the cut. You do not have to predict which state moves next to see that the direction is set.
For a multi-state employer, that means the practical question is not whether you operate in Connecticut today. It is whether you want to build the notice habit once, calmly, or scramble to build it five times as five more states each pass their own version. The first option is cheaper and far less stressful in the middle of a real RIF.
This briefing stays on the layoff-notice angle. If you want the wider state-by-state map of AI employment rules, including Texas and Illinois, that is a separate read: see the Texas TRAIGA and multi-state map.
What to put in the RIF playbook now
You can prepare for this without a lawyer in the room, then confirm the details with one. The work is mostly about documentation and sequence, and it is the same work whether or not your state has acted yet.
- Decide and record the role AI played. Before any cut, write down honestly how AI was used. Did it rank roles, model the budget, or score performance, or did a human make the call with AI only in the background? The answer determines whether a notice duty is even in play.
- Keep the human decision visible. A reduction in force should still trace to a named human who owns it. AI can inform the analysis. It should not be the thing you point to when an employee asks who decided.
- Draft the notice language ahead of time. If AI shaped the decision, have a plain, accurate notice ready that says so without overstating or hiding it. Writing it in advance, off the clock, beats writing it under deadline pressure.
- Align it with your existing layoff process. The federal and state thresholds you already track do not go away. Treat AI notice as an added layer on the RIF checklist, not a replacement for anything.
- Brief your managers. The people delivering the news need to understand that how AI was used is now a documented fact, not an offhand explanation they can improvise in the room.
What good notice actually looks like
If a notice duty is in play, the quality of the notice matters as much as the fact of it. A bad notice is vague, defensive, or written to obscure what happened. A good one is short, plain, and honest about the role AI played, without dressing it up or hiding behind it. The goal is a worker who reads it and understands what occurred, not one who walks away more confused or more suspicious than before.
In practice that means avoiding two opposite mistakes. The first is overstating the machine's role to deflect responsibility, as if the software fired anyone. It did not. A human acted on an analysis, and the notice should reflect that a person made the decision. The second mistake is understating it to the point of misleading, burying a meaningful AI input under generic language about restructuring. Both undermine the trust the notice is supposed to preserve, and both can look worse later if the real process ever comes to light.
There is also a sequencing question worth settling in advance. Notice that lands at the same moment as the layoff, with no warning and no context, reads as colder than notice delivered as part of a clear, humane process. The law sets a floor. How you deliver on it is a culture decision, and the firms that handle reductions well treat the AI disclosure as one piece of an already-respectful conversation rather than a bureaucratic afterthought stapled to a hard day.
The honest caveat
A fair question is whether this discourages using AI in workforce planning at all. It should not. AI is genuinely useful for modeling scenarios, stress-testing a budget, and surfacing patterns a tired human would miss at the end of a hard quarter. None of that becomes illegal.
What changes is that you can no longer treat the AI step as invisible. The old habit of running an analysis quietly and presenting only the conclusion is the habit that gets you in trouble. The new discipline is to use the tool and keep an honest record of how you used it. A documented, defensible process protects the company and treats the affected workers with the basic respect of a straight answer.
There is also a quieter benefit. The same documentation that satisfies a notice rule is the documentation that protects you in any layoff dispute. Knowing exactly how the decision was made, and being able to show it, is good practice regardless of which state you sit in.
The skill under the rule
Every few months a new AI employment rule lands somewhere, and the response is always the same scramble. The teams that stay calm are not the ones who memorized this week's statute. They are the ones with a working method: where AI belongs in a people decision, where it does not, and how to keep a clean enough record that a new disclosure or notice rule is a small addition rather than an emergency.
That method is what survives every new bill and every new state. If you want the structured version built for HR and people-ops, The Leveraged HR Professional teaches it from the ground up, and the two minute course quiz will point you to the right starting place for your team. For the broader picture of what HR already owes on the hiring side, our companion briefing is a good next read.
Frequently Asked Questions
Does this Connecticut law apply to my company if we are not in Connecticut?
The notice requirement is a Connecticut law, so it directly binds employers covered by Connecticut's rules. The reason it matters more widely is that it is a clean, copyable template, and AI is now the leading cited reason for layoffs in 2026 per Forbes. A rule aimed at the most common cause of cuts tends to spread, so building the habit early is the prudent move for any multi-state employer.
What exactly triggers the notice duty?
The trigger is the cause of the cut, not the headcount. Under the Connecticut law that Governor Lamont signed in late May 2026, when an employer's use of AI causes a reduction in force, affected employees are owed notice. That is different from the threshold-and-timing rules HR already tracks, which are keyed to how many people and how much warning.
Does this mean we should stop using AI in workforce planning?
No. AI remains useful for modeling scenarios and budgets. What changes is that you can no longer treat the AI step as invisible. The practical answer is to use the tool and keep an honest record of how you used it, so that if notice is required you can give a clear, defensible account of the decision.
Is this briefing legal advice?
No. The Leveraged Years is an education company, not a law firm. This is a plain language explainer of a fast moving story, and the specifics of any AI employment law can change and vary by state. Treat it as background, and confirm anything that affects your firm's compliance or your actual RIF process with a qualified employment attorney.