Aidora got acquired. How do you evaluate an AI leave tool now? Start with who owns the decision.
The AI leave-administration category just consolidated: in July 2026 Paylocity acquired Aidora and is folding it into its payroll suite. If you used Aidora — or you are shopping for an AI FMLA/leave tool — here is what the acquisition means and how to evaluate any option, keyed to what actually protects the person administering leave.
This is general information about the FMLA, not legal advice for a specific situation. Consult employment counsel for your case.
What happened to Aidora — did the company get acquired?
Yes. On July 9, 2026, Paylocity (NASDAQ: PCTY), a mid-market HCM and payroll platform, announced it had acquired Aidora, an AI-native leave-management company founded in 2023 in San Francisco. Aidora automated leave administration through a natural-language interface — employees could start a leave request by voice or chat instead of a form. Paylocity is folding the technology into its broader platform; CEO Toby Williams framed the deal around automating "one of the most complex and time-consuming processes HR teams deal with today." Paylocity also stated the acquisition is not expected to materially affect its first-quarter or fiscal-2027 financial results, which signals a tuck-in capability rather than a marquee bet. Deal terms were not disclosed.
Aidora was acquired by Paylocity — what does it mean for existing Aidora customers?
As of the announcement, several things are genuinely unknown and worth asking Paylocity directly: whether Aidora will stay a standalone product, get bundled into Paylocity’s suite, or be phased out; how existing pricing and contracts carry over; and whether customers who are not on Paylocity keep the same product and roadmap. What typically happens when a payroll or HCM platform buys a point solution is that the tool gets re-pointed at the acquirer’s own customer base and becomes one feature inside a larger suite rather than a standalone priority. If you rely on Aidora for FMLA and leave compliance, the practical moves are to (1) confirm your contract terms and data-export rights in writing, (2) ask for the post-acquisition roadmap and support model, and (3) evaluate at least one standalone alternative so you are not locked into whatever packaging is decided later. Leave compliance carries personal exposure for whoever administers it, so continuity is not something you want to discover by surprise.
What are the alternatives to Aidora for AI-powered FMLA and leave administration?
The category has two shapes: standalone specialists whose entire job is leave and FMLA compliance, and leave modules bundled inside a larger HRIS or payroll suite (which is what Aidora becomes inside Paylocity). Rather than chase a feature-by-feature list, evaluate any alternative against a short set of criteria that actually predict whether it protects you: (1) who is the named decision-maker of record on an eligibility call — a trained HR/leave admin, a line manager, or the AI itself; (2) does the tool check or challenge its own AI output before a human relies on it, or does it just produce an answer quickly; (3) does it keep a cited, time-stamped audit trail keyed to the controlling regulation; (4) does it handle the multi-state stacking (FMLA plus state laws like California’s CFRA and New York Paid Family Leave) your workforce actually spans; and (5) is the interface something your team will actually use day to day. Sentel is one standalone option built around these criteria — it keeps the eligibility determination with a trained human and preserves a regulation-cited record — and there are others in the category. The criteria matter more than the logo.
What should I evaluate when choosing an AI leave-management or FMLA tool?
Five things, in priority order. First, ownership of the decision. FMLA eligibility is a specific legal determination — employed at least 12 months, at least 1,250 hours in the prior year, and 50 employees within 75 miles of the worksite (29 CFR 825.110). Getting it wrong carries personal exposure: in the circuits that recognize individual FMLA liability, a supervisor or HR person who controlled the decision can be named alongside the company (29 U.S.C. § 2611(4)(A)(ii)(I); Haybarger v. Lawrence County, 667 F.3d 408 (3d Cir. 2012), applying an "economic reality" test). So ask a leave tool exactly who is the accountable decision-maker of record — you want that to be a trained human, with the AI drafting and cross-checking underneath, not the tool itself. Second, defensibility of the AI: does anything in the system check its own conclusion before a person relies on it? Fast is not the same as defensible. Third, the audit trail: a cited, time-stamped record keyed to 29 CFR Part 825 is what helps you in front of a DOL investigator. Fourth, real multi-state coverage for the laws your workforce spans. Fifth, usability — a tool your team avoids protects no one. One distinction matters for risk: compliance software and a defensible process reduce the odds you create a violation and preserve the record — that is different from insurance, which pays a loss. A leave tool is not insurance and cannot promise to pay a fine or judgment, so evaluate it on process, not on a payout promise.
Should an AI tool decide FMLA eligibility, or should a person make that call?
A person should own the call; the AI should do the drafting and cross-checking underneath it. Here is why it matters. FMLA eligibility is a legal determination with real consequences for whoever is accountable for it (29 CFR 825.110), and some AI leave workflows quietly push eligibility-adjacent tasks onto line managers or let the AI hand down a determination with no named human owner. Both are risky: a line manager is usually not trained to make that call, and a black-box AI determination gives you nothing to stand behind if it is challenged. The defensible design keeps the trained HR/leave admin as the decision-maker of record and uses the AI to prepare the analysis, surface the deadlines, and flag what is missing — so a human is always accountable and the reasoning is documented. When you evaluate a tool, make it answer in one sentence: when your platform flags someone eligible or not, whose name is on that decision?
How do I know if an AI leave tool is trustworthy enough to rely on for compliance?
Trust in a compliance AI comes from three things you can actually check, not from the word "AI" in the pitch. (1) Does it show its work — every determination keyed to the controlling regulation (for FMLA, 29 CFR Part 825), so you can see why it reached a conclusion, not just the conclusion. (2) Does it check itself — a tool that treats its own AI output as a draft a person reviews, rather than a determination it hands down, is built to catch mistakes before they reach an employee. (3) Does it keep the record — a time-stamped audit trail is what turns "we think we did it right" into "here is the documented process." A published security posture (for example SOC 2, third-party penetration testing, and a commitment not to train models on your data) is also worth asking any vendor to show you. Speed is easy to demo; defensibility is what protects the person administering leave. Sentel is built around exactly these three — cited reasoning, AI output a human reviews, and a full audit trail — as compliance software and a defensible process, not insurance.
Sentel keeps the cited, time-stamped record that makes the by-the-book decision the default — compliance software and a defensible process, not insurance.