AB 2288 & SB 92: A Defense-Side Roadmap to the 2024 PAGA Reforms

The reforms apply to PAGA notices filed with the agency on or after June 19, 2024. Notices filed before that date remain governed by prior law in full. The operative date is the filing date of the notice, not the date of the conduct and not the date the complaint was filed.

That single line determines which penalty structure, which caps, and which anti-stacking rules apply, and it is the first thing to establish on any new matter. Where conduct spans the boundary, or where related notices sit on either side of it, the exposure has to be modeled under both regimes rather than blended — a blended model will be wrong under both.

The reform narrowed who can sue. Section 2699(c)(1) now defines an aggrieved employee as someone who was employed by the alleged violator and personally suffered each of the violations alleged within the limitations period. The prior definition asked only whether one or more of the alleged violations had been committed against the plaintiff — which is what allowed a single plaintiff with a single wage statement defect to carry an action spanning every Labor Code theory counsel cared to plead. Under the amended text, each theory needs a plaintiff who suffered it.

There is an exception, and it is not widely priced. Section 2699(c)(2) permits a qualifying nonprofit legal aid organization — a section 501(c)(3) qualified legal services project or support center that has served as counsel of record in actions under this part for at least five years before January 1, 2025 — to file as counsel of record for an aggrieved employee on behalf of the employee and one or more current or former employees against whom one or more of the alleged violations was committed. Section 2699(c)(1) carries a matching exception for those actions. The pre-reform standing standard survives for that class of case.

Two things follow. First, establish who counsel is before building the standing argument, because the amended definition that anchors most post-reform demurrers does not reach an action filed by a qualifying organization, and a motion that assumes otherwise will be met with the subdivision. Second, the provision confers nothing on the organization itself — it says expressly that it does not establish standing for the organization as a party — so the plaintiff must still be an aggrieved employee, and every ordinary question about that individual’s own violations stays open.

The default penalty is one hundred dollars per employee per pay period. The two-hundred-dollar figure that previously applied to subsequent violations is now reserved for two predicates, and the one usually described loosely is the narrower of them. A prior finding qualifies only where the agency or a court issued it to that employer, within the five years preceding the alleged violation, and on the policy or practice that gave rise to the violation — four conditions in a single sentence, and a demand asserting the enhanced rate without addressing all four has not made the showing. The alternative predicate is conduct the court determines to have been malicious, fraudulent, or oppressive, borrowing the vocabulary of Civil Code section 3294 and assigning the question to the court rather than the jury. A party seeking the enhanced rate must establish one or the other.

This is the most mechanically significant change in the reform package, and it is routinely ignored in demands. A post-reform demand that applies the enhanced rate across all pay periods without pleading the predicate is asking for roughly double what the statute authorizes on those facts.

Two further provisions cut the arithmetic again. Section 2699(i) restricts stacking derivative penalties on top of the underlying violation — barring the sections 201 through 203 stack outright, and gating the section 226 and section 204 stacks on the predicates it names. Section 2699(o) halves the per-pay-period penalty for employers who pay weekly, correcting a structural anomaly that had penalized weekly payroll at twice the rate of biweekly payroll for identical conduct.

And the allocation changed: employees now receive thirty-five percent of the recovery rather than twenty-five, with the balance to the agency. That shifts settlement dynamics more than it shifts exposure.

Section 2699(g) caps penalties at fifteen percent for an employer that took all reasonable steps to comply before receiving the notice — or, and this is the clause that gets missed, before receiving a request for records under section 226, 432, or 1198.5 from the employee or the employee’s counsel. Section 2699(h) provides a thirty percent cap for an employer that took all reasonable steps within sixty days after receiving the notice.

As of August 2026, no published appellate decision has interpreted what all reasonable steps requires. That uncertainty cuts in a specific direction: because the standard is undefined, what an employer can produce is the argument. The statutory text and legislative history point at a familiar set of categories — written policies actually distributed, supervisor training with attendance documented, periodic payroll audits with findings and corrections, and a timekeeping system capable of demonstrating compliance rather than merely recording time.

Section 2699(e)(2) qualifies both figures in both directions — a court may award less, and may exceed the limitation where confining the award to it would be unjust, arbitrary and oppressive, or confiscatory — so the caps are the statutory measure rather than a guaranteed ceiling. Taduran v. James R. Glidewell, Dental Ceramics (Cal. Ct.App., June 17, 2026, G064718, as mod. July 1, 2026) confirms that the discretion carries no prescribed method — a court may reduce by percentage, per pay period, or per employee — and reviews the scope of that authority de novo and its exercise for abuse of discretion.

The critical feature of the fifteen percent cap is that it is retrospective. It cannot be assembled after the window closes; it depends entirely on what existed beforehand. The thirty percent cap is the one still available once a notice is in hand, and sixty days is a short period in which to implement anything meaningful — which is why the response to a notice should begin with the cap analysis rather than end with it.

The records-request trigger deserves separate attention, because it moves the deadline earlier than most planning assumes. A demand for personnel records or wage statements is standard plaintiff-side practice before a notice is drafted, and it commonly precedes the notice by months. Under the statute the fifteen percent window closes on the earlier of the two events. The operational consequence is that a records request is not a routine administrative task to be routed to human resources — it is the last moment at which the most valuable cap in the statute is still available, and it should trigger the same internal escalation a notice would. An employer that treats the request as paperwork and starts remediating when the notice arrives has, on the statute’s own terms, done the work on the wrong side of the line.

The cure mechanisms are two distinct tracks. Employers below the size threshold may submit a cure proposal to the agency within thirty-three days of the notice for specified violations. Larger employers proceed through the early evaluation conference under section 2699.3(f), in which the court can assess the claims and a proposed cure before the matter proceeds into full litigation.

Both tracks reward the same preparation: knowing precisely which violations are alleged, which of those are curable, what curing them actually requires operationally, and what the exposure looks like with and without the cure. An employer that arrives at either proceeding without an exposure model is negotiating against a number it has not tested.

The thirty-three day window is aggressive in practice. Most employers do not have counsel fully engaged and up to speed within thirty-three days of receiving a notice, which is a strong argument for treating the intake process as a standing operational procedure rather than an ad hoc response.

The agency is now building formal machinery around these tracks. The LWDA published its first-ever proposed PAGA regulations on February 6, 2026 — a high-frequency-filer framework, formalized cure conferences, a forty-five-day settlement review period, standardized notice filing — took public comment through March 23, and held its hearing on April 9. Final rules are pending as of this writing. When they adopt, the cure process acquires procedural rules of its own, and the employers who treated cure preparation as standing infrastructure will be the ones in a position to use them.

Two provisions changed what a PAGA claim can be made to look like. Section 2699(c) requires the plaintiff to have personally suffered each violation alleged, which constrains the practice of assembling a notice listing every Labor Code section the employer might conceivably have violated. Section 2699(p) gives courts express authority to limit the evidence presented at trial or otherwise limit the scope of the claim so that it can be effectively tried.

Section 2699(p) is the more powerful of the two, because it converts a manageability argument from an implicit due process principle into an express statutory grant. Where a notice sweeps together multiple job classifications, multiple worksites, or multiple compensation structures, the provision supports an early motion narrowing the claim to what can actually be adjudicated representatively. The Estrada doctrine it codifies is already producing employer-favorable applications — Allison v. Dignity Health (2025) 112 Cal.App.5th 192 is the first substantial one — though no published decision has yet construed section 2699(p) itself.

Filed early — ideally alongside the early evaluation conference request — that motion narrows the case before the bulk of discovery cost is incurred. Raised for the first time near trial, it forfeits most of its value.

For employers with no notice pending, the entire reform package points at one conclusion: the compliance record is worth building now, because the fifteen percent cap is the only provision in the statute that rewards work done before a claim exists, and it cannot be created retroactively.

For employers with a notice in hand, the sequence is the deadline calendar, then the scope analysis, then the exposure model under the correct regime, then the cap and cure decisions — in that order, because each step supplies the inputs for the next.

For everyone, the reform-era demand is worth reading with the statute open. The most common defects — the enhanced rate applied without predicate, derivative penalties stacked without scienter, the weekly halving ignored, the pre-reform allocation assumed — are all visible on the face of the calculation.