[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fGhCiwD2pXmbvp-0bTeT_qM_qXJtWRdCPQNhjGcAxF-c":3,"$fm6Ky0r9HHO8PTXKBgXJeU4fCLm4z6RKRMXFRXG3f560":28},{"success":4,"data":5},true,{"id":6,"slug":7,"title":8,"excerpt":9,"content":10,"category":11,"tags":12,"author":20,"cover_image_url":21,"reading_time_minutes":22,"is_published":4,"published_at":23,"created_at":24,"updated_at":24,"author_avatar":21,"is_featured":25,"meta_title":26,"meta_description":27,"meta_keywords":21},"73e63698-b636-41e1-8c33-98178f10f0af","no-robo-bosses-act-human-in-the-loop-2026","California Just Told Employers: A Human Has to Actually Decide","California's new No Robo Bosses Act bars AI-only firing and discipline decisions. Here is what it actually requires, why an earlier version was vetoed, how it differs from NYC and Illinois AI-hiring law, and what any employer running automated hiring or performance tools should check now.","\u003Cp>On September 30, 2026, California Governor Gavin Newsom signed SB 947 — the \"No Robo Bosses Act\" — making California the first U.S. state to bar employers from disciplining or firing a worker based solely on an automated system's output (\u003Ca href=\"https:\u002F\u002Fnews.bloomberglaw.com\u002Fdaily-labor-report\u002Fnew-california-law-requires-that-humans-decide-firings-not-ai\">Bloomberg Law\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fupi.com\u002FTop_News\u002FUS\u002F2026\u002F10\u002F01\u002Fcalifornia-artificial-intelligence\u002F5411790881596\">UPI\u003C\u002Fa>). The law doesn't ban AI from employment decisions. It bans AI from being the \u003Cem>whole\u003C\u002Fem> decision. That distinction — assist versus replace — is quietly becoming the organizing idea behind a growing body of state and local law, and it has direct implications for anyone running hiring or performance tools, not just the discipline-and-termination use case the new California law actually targets.\u003C\u002Fp>\n\n\u003Ch2>What SB 947 actually requires\u003C\u002Fh2>\n\n\u003Cp>SB 947 defines an \"automated decision system\" broadly: any computational process using machine learning, statistical modeling, data analytics, or AI that produces a score, classification, or recommendation used to assist or replace human discretionary decision-making and materially affects a person (\u003Ca href=\"https:\u002F\u002Fwww.fmglaw.com\u002Femployment\u002Fcan-ai-be-the-boss-california-bill-targets-automated-workplace-decisions\u002F\">Freeman Mathis & Gary\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fhoodline.com\u002F2026\u002F09\u002Fcalifornia-bill-would-stop-bosses-from-letting-ai-fire-workers-solo\u002F\">hoodline\u003C\u002Fa>). Under the law:\u003C\u002Fp>\n\n\u003Cul>\n\u003Cli>An employer cannot use such a system as the \u003Cstrong>sole basis\u003C\u002Fstrong>, or as the \u003Cstrong>principal tool\u003C\u002Fstrong>, for a disciplinary or termination decision.\u003C\u002Fli>\n\u003Cli>Where an employer relies primarily on the system's output, a \u003Cstrong>human must corroborate\u003C\u002Fstrong> the decision using independent inputs — the bill names managerial evaluations, peer reviews, and personnel files as examples — and the output cannot be used if it's inaccurate, incomplete, or misleading (\u003Ca href=\"https:\u002F\u002Fwww.cdflaborlaw.com\u002Fblog\u002Frobots-can-recommend-but-real-people-must-pull-the-trigger-inside-sb-947\">CDF Labor Law\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fsynhr.com\u002Fsb-947-california-ai-discipline-termination\u002F\">synhr\u003C\u002Fa>).\u003C\u002Fli>\n\u003Cli>A worker disciplined or terminated under a system that primarily relied on AI must receive \u003Cstrong>written notice\u003C\u002Fstrong>: that AI was involved, what data it drew on, and a human contact who can explain the decision.\u003C\u002Fli>\n\u003C\u002Ful>\n\n\u003Cp>The law becomes operative July 1, 2027 — giving employers about nine months from today to build the corroboration workflow and the paper trail it requires, not the day the fine first arrives.\u003C\u002Fp>\n\n\u003Ch2>The bill that failed first, and why that matters\u003C\u002Fh2>\n\n\u003Cp>The more interesting part of this story, for anyone trying to predict where regulation goes next, is what happened the year before. Newsom vetoed an earlier version of this bill — SB 7 — in October 2025, calling it \"unfocused\" and \"overly broad\" (\u003Ca href=\"https:\u002F\u002Fwww.hrdive.com\u002Fnews\u002Fgov-newsom-vetoes-no-robo-bosses-act\u002F802758\u002F\">HR Dive\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fwww.fisherphillips.com\u002Fen\u002Finsights\u002Finsights\u002Fcalifornia-governor-vetoes-no-robo-bosses-act.html\">Fisher Phillips\u003C\u002Fa>). State Senator Jerry McNerney reintroduced it in 2026 as SB 947 with two things stripped out: a pre-notification requirement and provisions covering gig workers, which had drawn direct opposition from Uber and Lyft (\u003Ca href=\"https:\u002F\u002Fnews.bloomberglaw.com\u002Fdaily-labor-report\u002Fnew-california-law-requires-that-humans-decide-firings-not-ai\">Bloomberg Law\u003C\u002Fa>).\u003C\u002Fp>\n\n\u003Cp>The lesson isn't \"California tried twice and got lucky.\" It's that the core idea — a human has to be able to explain and correct an automated decision, not just rubber-stamp it — survived a veto that killed everything built around it. That's usually a sign a principle is going to keep reappearing in other states' bills even where the specific mechanics (notice timing, which workers are covered) get rewritten or dropped.\u003C\u002Fp>\n\n\u003Ch2>Three different regulatory models, not one trend\u003C\u002Fh2>\n\n\u003Cp>It's tempting to read SB 947 as part of one unified \"AI hiring law\" story. It isn't — and treating three distinct regulatory models as interchangeable is exactly how compliance gaps happen for employers operating in more than one jurisdiction.\u003C\u002Fp>\n\n\u003Cp>\u003Cstrong>New York City's Local Law 144\u003C\u002Fstrong> (effective January 2023, enforced since July 2023) is a disclosure-and-audit regime aimed specifically at hiring and promotion tools. It requires an independent bias audit no more than a year old, public posting of a summary of that audit's results, and advance notice to candidates — enforced by the Department of Consumer and Worker Protection with civil penalties of $500–$1,500 per violation per day (\u003Ca href=\"https:\u002F\u002Frules.cityofnewyork.us\u002Frule\u002Fautomated-employment-decision-tools-2\u002F\">NYC Rules\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fbabl.ai\u002Fai-audits\u002Fnyc-bias-audit\u002F\">babl.ai\u003C\u002Fa>). It says nothing about who makes the final call — it says you have to measure and disclose what the tool does.\u003C\u002Fp>\n\n\u003Cp>\u003Cstrong>Illinois's HB 3773\u003C\u002Fstrong>, effective January 1, 2026, folds AI use directly into the Illinois Human Rights Act, covering recruitment, hiring, discharge, and discipline. It requires notice when AI is used in an employment decision and bars using a zip code as a proxy for a protected class — but, notably, offers no audit safe harbor the way NYC's law effectively does. A discriminatory \u003Cem>outcome\u003C\u002Fem> is the violation regardless of what audit the employer ran beforehand (\u003Ca href=\"https:\u002F\u002Fogletree.com\u002Finsights-resources\u002Fblog-posts\u002Fillinois-steps-up-ai-regulation-in-employment-key-takeaways-for-employers\u002F\">Ogletree Deakins\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fvensure.com\u002Femployment-law-updates\u002Freminder-to-illinois-employers-provide-ai-notice-starting-january-1-2026\u002F\">Vensure\u003C\u002Fa>).\u003C\u002Fp>\n\n\u003Cp>\u003Cstrong>California's SB 947\u003C\u002Fstrong> is neither a disclosure regime nor an outcomes regime — it's a \u003Cem>process\u003C\u002Fem> mandate. It doesn't require an audit, and it doesn't make a bad outcome the violation on its own. It requires a specific workflow: AI can recommend, but a human has to independently corroborate before anyone acts, and that corroboration has to be documented.\u003C\u002Fp>\n\n\u003Cp>Three different compliance obligations, three different things to build, three different ways to fail an audit. An employer who satisfies NYC's disclosure requirement has done nothing toward California's corroboration requirement, and vice versa. Multi-state employers using the same AI-assisted screening or performance tool across offices in New York City, Illinois, and California are, as of today, subject to all three regimes at once for the same tool, depending on where the affected worker sits.\u003C\u002Fp>\n\n\u003Cp>It's also worth being precise about scope, because the three laws don't cover the same stage of employment. NYC's and Illinois's laws reach into \u003Cem>hiring\u003C\u002Fem> — screening and selecting candidates. California's SB 947, as written, is about \u003Cem>discipline and termination\u003C\u002Fem> of people who are already employed; it doesn't regulate the hiring decision itself. A tool that only screens incoming applicants and never touches a disciplinary file sits outside SB 947's current scope even while it sits squarely inside NYC's and Illinois's. That distinction matters for anyone trying to map which law applies to which part of their stack — and it's also the detail easiest to get wrong if you only read the headline.\u003C\u002Fp>\n\n\u003Ch2>The number of these laws is genuinely contested — and that's itself informative\u003C\u002Fh2>\n\n\u003Cp>If you go looking for \"how many states regulate AI in hiring,\" you'll get a different number depending on which tracker you read: some count roughly five or six states with direct statutory obligations, others count a dozen or more once city ordinances, narrower sectoral rules, and employment-specific provisions inside broader state AI acts are folded in. Colorado's own AI Act illustrates why counting is hard: it was enacted in 2024, delayed to mid-2026, then repealed and re-enacted as a narrower disclosure law taking effect January 2027 — a law that existed, didn't take effect, and got replaced by a different law before it ever bound anyone. Any tracker's count is a snapshot of a landscape that is still actively being rewritten, not a stable fact to cite once and reuse.\u003C\u002Fp>\n\n\u003Cp>What isn't contested, across every tracker and every framing, is the direction: there is no state currently moving toward \u003Cem>less\u003C\u002Fem> obligation on employers using AI in employment decisions, and the newest entrant (California) chose the strictest of the three models — a mandatory human-corroboration workflow — not the lighter disclosure-only approach NYC pioneered four years earlier.\u003C\u002Fp>\n\n\u003Ch2>What this means if you're the one deciding\u003C\u002Fh2>\n\n\u003Col>\n\u003Cli>\u003Cstrong>Find every auto-decision toggle in your stack and check which way it's set, today, not in nine months.\u003C\u002Fstrong> If any tool in your hiring or performance pipeline has a setting that lets an automated score or classification act without a human reviewing it, you now have a California compliance deadline attached to that setting even if you never thought of it as one. \"We've always had that option\" is not the same as \"we've documented that we're not using it.\"\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Build the corroboration record before the deadline forces it.\u003C\u002Fstrong> SB 947's human-review requirement is specific about what counts: managerial evaluations, peer reviews, personnel files — not a manager glancing at a dashboard and clicking approve. If your current process for acting on an automated recommendation doesn't produce a record of independent human input, July 1, 2027 is the date that gap becomes a liability rather than a process footnote.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Don't assume one compliance project covers every state you operate in.\u003C\u002Fstrong> A bias audit that satisfies NYC's Local Law 144 does nothing for California's corroboration requirement, and a human-review workflow built for California doesn't discharge Illinois's notice obligation or its zip-code-proxy rule. Map tool, use case, and jurisdiction separately rather than building one \"AI compliance\" checklist and assuming it travels.\u003C\u002Fli>\n\u003C\u002Fol>\n\n\u003Cp>On our own platform, the default has deliberately been on the side this new wave of law is pushing toward, since well before SB 947 existed. A below-threshold candidate in NiceHire's resume vetting or AI screening stages routes to pending review for a human decision by default; auto-reject without human review is available, but it's an opt-in setting an organization has to explicitly turn on per stage, not the out-of-the-box behavior. Every candidate who reaches AI screening is scored against the same fixed criteria — technical, communication, and cultural fit, each on a 0–100 scale — against a threshold the employer sets in advance, which is a statement about a consistent mechanism, not a claim about fairness or accuracy. We note this not because our tool is covered by SB 947 as written — it screens candidates into a hiring pipeline, which sits outside that law's discipline-and-termination scope, not inside it — but because the underlying design question is the same one SB 947 answers for firings: does an automated score get to be the whole decision, or does it have to be a recommendation a human can check? For how we answer that question on our own screening stages, the law and the product default already point the same way.\u003C\u002Fp>\n\n\u003Cp>That alignment isn't a compliance guarantee, and it isn't ours to claim as one. It's a reason to go check your own stack's defaults now, while the deadline is still nine months out instead of nine days.\u003C\u002Fp>","Compliance",[13,14,15,16,17,18,19],"AI regulation","California SB 947","No Robo Bosses Act","AI hiring law","compliance","HR tech","automated decision systems","NiceHire Team",null,8,"2026-10-04T00:27:46.583+00:00","2026-10-04T00:27:47.117317+00:00",false,"No Robo Bosses Act: What California SB 947 Means for Hiring Tools","California SB 947 bars AI-only firing and discipline decisions starting July 2027. How it compares to NYC Local Law 144 and Illinois HB 3773, and what to check in your stack now.",{"success":4,"data":29},{"posts":30,"count":68,"hasMore":4},[31,33,45,58],{"id":6,"slug":7,"title":8,"excerpt":9,"category":11,"tags":32,"author":20,"cover_image_url":21,"reading_time_minutes":22,"published_at":23},[13,14,15,16,17,18,19],{"id":34,"slug":35,"title":36,"excerpt":37,"category":11,"tags":38,"author":20,"cover_image_url":21,"reading_time_minutes":22,"published_at":44},"e9054a4f-74e8-4d0c-97a9-bcb10a585447","illinoiss-ai-hiring-law-took-effect-in-january-its-own-compliance-rules-didnt","Illinois's AI Hiring Law Took Effect in January. Its Own Compliance Rules Didn't.","HB 3773 made AI-caused hiring discrimination a civil rights violation on January 1, 2026. The notice rules meant to define compliance were withdrawn in June and never refiled. Here's what's actually binding right now, and what isn't.",[16,39,40,41,42,13,43],"Illinois HB 3773","employment discrimination","HR compliance","disparate impact","hiring trends","2026-10-03T00:28:37.906+00:00",{"id":46,"slug":47,"title":48,"excerpt":49,"category":11,"tags":50,"author":20,"cover_image_url":21,"reading_time_minutes":56,"published_at":57},"3e54cabb-8ad7-468c-aa08-a600613d011c","ai-hiring-vendor-liability-2026","The Hiring Software Is Now a Legal Party to the Decision","A federal court let disparate-impact claims against Workday proceed on an agent theory, Illinois made undisclosed AI use its own civil rights violation, and a California bill requiring human corroboration sits on the governor's desk today. Three separate mechanisms, one direction: the software is now a legal party to the hiring decision.",[51,52,53,54,39,14,41,55],"AI hiring","employment law","vendor liability","Mobley v Workday","algorithmic discrimination",9,"2026-09-30T00:25:30.211+00:00",{"id":59,"slug":60,"title":61,"excerpt":62,"category":11,"tags":63,"author":20,"cover_image_url":21,"reading_time_minutes":56,"published_at":67},"6e431f94-fecc-44be-aff1-0fbbcb99c511","mobley-v-workday-ai-hiring-vendor-liability","The Workday Lawsuit Isn't About Workday: What Three Years of Mobley v. Workday Rulings Mean for Every AI Hiring Vendor","Mobley v. Workday has spent three years and six rulings deciding whether an AI hiring vendor — not just the employer using its tool — can be sued directly for discrimination. Here's what's actually settled, what's still moving, and what it means for anyone building or buying algorithmic screening tools.",[51,52,64,65,41,53,66],"Mobley v. Workday","algorithmic bias","AI Act","2026-09-13T00:19:17.41+00:00",20]