Regulatory Exposure Intelligence — sample brief

A large enterprise — named, because the record is public

Generated 2026-08-23 · engine 2.0.0-phase2 · 246 matters held in the corpus at generation

RealPage, Inc. is named here rather than masked: every matter behind this brief is a public enforcement record, and a sample nobody can check is not evidence of anything. The profile below (sector, size, states, AI applications) describes the company as it is publicly known; the FINDINGS are whatever the engine computed from the corpus, unedited.

AI regulatory data brief — RealPage, Inc.

Generated 23 August 2026 · engine 2.0.0-phase2 · deterministic rules over retrieved records; no model in this path

Automated analysis of public legal data. This states what the data shows; it is not legal advice, not an assessment of whether any obligation has been met, and not a prediction of enforcement. Every finding links to its source. Decisions should be taken with qualified counsel.

Profile this brief was run against

At a glance

1. What reaches this company

In force (6)

• SB 942 — AI Transparency Act
  California · in effect from 2 August 2026
  Generative-AI providers with over 1,000,000 monthly users must offer a free AI-detection tool and embed a latent provenance disclosure in AI-generated image, video and audio content, plus an optional visible manifest disclosure. Operative August 2, 2026 (delayed from January 1, 2026 by AB 853).
  Penalty provided by the instrument: $5,000 per violation; each day is a discrete violation

⚖ This instrument's text contains a scale threshold — "over 1,000,000 monthly users". The threshold is reproduced as written; whether this company falls inside it is a question of its own numbers and is not determined here.

From the instrument's own text, read from the official source:

SB 942(a) — “A covered provider shall offer the user the option to include a manifest disclosure in image, video, or audio content, or content that is any combination thereof, created or altered by the covered provider’s GenAI system that meets all of the following criteria:” SB 942(b) — “A covered provider shall include a latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered provider’s GenAI system that meets all of the following criteria:” SB 942(c)(1) — “If a covered provider licenses its GenAI system to a third party, the covered provider shall require by contract that the licensee maintain the system’s capability to include a disclosure required by subdivision”

Quoted from the instrument's text — wording unchanged, whitespace normalised. 3 of 62 provisions in the document matched; read the full instrument at the link above.

• TRAIGA — Texas Responsible AI Governance Act (HB 149, 2025)
  Texas · in effect from 1 January 2026
  Prohibits developing or deploying AI for intentional behavioral manipulation causing harm, unlawful discrimination, and unlawful synthetic media; applies to businesses and state agencies. Enforced exclusively by the Texas Attorney General with a 60-day cure period.
  Penalty provided by the instrument: AG-enforced (no private right of action); up to $100,000 per uncurable violation + $40,000/day
  From the instrument's own text, read from the official source:

TRAIGA Sec. 552.051(f) — “If an artificial intelligence system is used in relation to health care service or treatment, the provider of the service or treatment shall provide the disclosure under Subsection” TRAIGA Sec. 552.054(b) — “A governmental entity may not develop or deploy an artificial intelligence system for the purpose of uniquely identifying a specific individual using biometric data or the targeted or untargeted gathering of images or other media from the Internet or any other publicly available source without the individual's consent, if the gathering would infringe on any right of the individual under the United States Constitution, the Texas Constitution, or state or federal law.” TRAIGA Sec. 552.051(b) — “A governmental agency that makes available an artificial intelligence system intended to interact with consumers shall disclose to each consumer, before or at the time of interaction, that the consumer is interacting with an artificial intelligence system.”

Quoted from the instrument's text — wording unchanged, whitespace normalised. 3 of 217 provisions in the document matched; read the full text at https://capitol.texas.gov/tlodocs/89R/billtext/html/HB00149F.htm.

• Equal Credit Opportunity Act (ECOA)
  United States (federal) · 15 U.S.C. § 1691 · Federal Reserve; Consumer Financial Protection Bureau · in effect from 28 October 1974
  AI mortgage and home equity loan systems cannot discriminate. Requires disparate impact analysis.
  Penalty provided by the instrument: Up to $100,000 per violation; individual and Class action damages
  The instrument's text was read at analysis time — The source was read, but no provision in it both imposed a duty and mentioned the subject matter of this profile. See the source above for the full text.
• Fair Credit Reporting Act (FCRA) § 1681
  United States (federal) · 15 U.S.C. § 1681 · Federal Trade Commission; Consumer Financial Protection Bureau · in effect from 26 October 1970
  AI credit and background check systems used in rental decisions must be transparent and non-discriminatory.
  Penalty provided by the instrument: Actual damages or $100–$1,000 per violation; Class action liability
  The instrument's text could not be read at analysis time: source returned HTTP 404. Nothing about its contents is asserted here — see the source above.
• Fair Housing Act (FHA)
  United States (federal) · 42 U.S.C. § 3601-3619 · Department of Housing and Urban Development (HUD) · in effect from 11 April 1968
  AI-based property valuations, lending decisions, and rental screening cannot discriminate based on protected classes (race, color, national origin, religion, sex, disability, familial status).
  Penalty provided by the instrument: Civil penalties up to $30,000 (first violation); up to $80,000 (subsequent); damages; injunctive relief
  The instrument's text could not be read at analysis time: source returned HTTP 403. Nothing about its contents is asserted here — see the source above.
• HUD AI Guidance (2023)
  United States (federal) · HUD-funded research and guidance on fair lending · Department of Housing and Urban Development · no commencement date recorded in the corpus
  Recommends AI bias testing, explainability, and human oversight in property valuation and lending. Warns against redlining through algorithmic discrimination.
  Penalty provided by the instrument: HUD enforcement; compliance reviews; potential funding restrictions
  The instrument's text could not be read at analysis time: source returned HTTP 403. Nothing about its contents is asserted here — see the source above.

Enacted, not yet in force (1)

• SB 24-205 — Colorado AI Act (amended 2026 by SB 26-189)
  Colorado · in effect from 1 January 2027
  The most comprehensive US state AI law. As amended by SB 26-189 (2026) it takes effect January 1, 2027 and centers on transparency/disclosure for consequential automated decisions (the original algorithmic-discrimination duty of care was repealed). Attorney General rulemaking is underway; no final rules have been published yet.
  Penalty provided by the instrument: AG-enforced (Colorado Consumer Protection Act); up to ~$20,000 per violation
  The instrument's text was read at analysis time — The source was read, but no provision in it both imposed a duty and mentioned the subject matter of this profile. See the source above for the full text.

States with no AI statute in the corpus

• Washington: No comprehensive AI law — high-risk AI bill (HB 2157) died in committee; narrow measures only (companion chatbots, HB 2225; AI content disclosure, HB 1170)

Confidence: HIGH — High — multiple records, primary sources

Basis: Rule-based resolution over the curated statute corpora. Every instrument is retrieved, never generated.

2. Dates on the record

Ahead

Already in effect (5)

Commencement dates that have passed. This records when each instrument came into effect; it is not a statement about anything this company has or has not done.

Applicable, no provable date (1)

Listed rather than dated. A commencement date is only printed where the corpus carries one.

• HUD AI Guidance (2023) — the corpus carries no effective date for this instrument

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: Dates carried by finding 1. No date is asserted that the corpus does not carry or that the parser could not extract verbatim.

Limits of this finding:

• 1 applicable instrument(s) carry no provable commencement date and are listed separately rather than dated.

3. Enforcement activity in this area

Direction

RISING. 22 matched matters in the 12 months to 23 August 2026, against 15 in the 12 months before that (change +7, 46.7%).

Sources present in only one window

These lanes contributed to one comparison window and nothing to the other. Where that happens, part of the difference between the two windows is which sources this product had collected, not what regulators did.

11% of the compared matters come from these one-sided lanes.

Total matched, all time

96 matched matters. A total across the whole corpus, with no time window. It shows how much matched enforcement EXISTS; it is not evidence of a change in activity and must never be read as one.

Of these, 90 carry a provable date and 6 do not; the dated ones run from 16 July 2015 to 6 July 2026 across 17 distinct months.

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: Matter counts from the enforcement corpus, windowed on the earliest proven date per matter. A trend is only graded when the reliability gate passes.

Limits of this finding:

• 11% of the compared matters come from sources present in only one of the two windows (Texas Attorney General), so part of the difference reflects source coverage rather than enforcement activity.

4. Actions brought against comparably positioned companies

Monetary relief recorded

Covers the 5 matched matters that record a monetary amount, out of 96 matched. Matters recording no amount are not zero-valued; the corpus simply holds no figure for them. Duplicate rows are excluded, so no settlement is counted twice.

Kinds of relief imposed

Matters (25 of 96 shown)

#### When It Comes to Data Privacy, Consumers Must Be in the Driver’s Seat: Attorney General Bonta, Partners Secure $12.75 Million General Motors Privacy Settlement
CPPA · California · enforcement action · status: unknown · earliest recorded date 8 May 2026

OAKLAND, CA — California Attorney General Rob Bonta, together with San Francisco County District Attorney Brooke Jenkins, Los Angeles County District Attorney Nathan J. Hochman, Napa County District Attorney Allison Haley, and Sonoma County District Attorney Carla Rodriguez, and with support from the California Privacy Protection Agency (CalPrivacy), today announced a settlement with General Motors (GM) regarding its illegal sale of hundreds of thousands of Californians’ location and driving data to two data brokers in violation of the California Consumer Privacy Act (CCPA) and California’s Unfair Competition Law. The settlement, which is subject to court approval, includes $12.75 million in civil penalties and strong injunctive terms, including restrictions on its use of consumer driving data and a ban on such data being sold to data brokers. “General Motors sold the data of California drivers without their knowledge or consent and despite numerous statements reassuring drivers that it would not do so. This trove of information included precise and personal location data that could identify the everyday habits and movements of Californians,” said Attorney General Rob Bonta. “Today

Cited: California Consumer Privacy Act; CCPA; Delete Act

Relief: Monetary $12.75 Million

Matched on: ai_role, home_jurisdiction (AI application: algorithmic_pricing)

#### Attorney General Phil Weiser announces $7M settlement with corporate landlord LivCor for role in algorithmic rent-setting scheme
CO-AG · Colorado · enforcement action · status: unknown · earliest recorded date 18 June 2026

June 18, 2026 (DENVER) – Attorney General Phil Weiser today, as part of a bipartisan coalition of nine attorneys general, announced a $7 million settlement with LivCor, LLC, one of the property management companies named in ongoing antitrust litigation against software company RealPage. The settlement resolves allegations that LivCor used RealPage’s revenue management system to fix rental prices with competing landlords by illegally sharing and gathering confidential pricing information. This conduct interfered with the normal competitive process and enabled landlords to keep prices higher, even in conditions when landlords naturally would lower prices. Under today’s settlement (PDF), LivCor agrees to not use software offered by any company that uses competitively sensitive information to fix rent prices and agrees to cooperate in the ongoing prosecution of RealPage and other defendant landlords. “Coloradans are getting squeezed by the high cost of housing. It’s outrageous that corporate landlords use algorithms to collude, artificially hike rents, and maximize their profits off the backs of workers, seniors, families who are just looking for a place to live. Ending the use of algorithmic rental price fixing is just one of many tools we need to level the playing field. Today’s settlement will end this anticompetitive conduct and bring fairness back to the market for renters,” said Attorney General Weiser. RealPage uses algorithmic models to recommend price increases to subscribers. As alleged the January 2025 antitrust lawsuit, LivCor and other landlords, including five co-defendants, shared competitively sensitive data to generate pricing recommendations using RealPage’s algorithms. LivCor and other landlords discussed competitively sensitive topics — including pricing strategies, rents, and selected parameters for RealPage’s software — directly with each other. RealPage knew what competing landlords were charging and could increase profits for landlords by using that nonpublic information to recommend landlords set or raise their prices uniformly, thereby eliminating competition. This scheme left renters with no choice but to pay artificially high prices. As of February 2025, LivCor managed 10 properties in Colorado totaling 3,352 units that used RealPage’s pricing software. Today’s settlement, subject to court approval, requires LivCor to pay $7 million in penalties and fees to the states. LivCor must also: Cease use of any revenue management software that uses competitors’ nonpublic pricing data to generate rent recommendations. LivCor has stopped using RealPage software. Refrain from sharing competitively sensitive pricing information with rival landlords or property managers. Establish an antitrust compliance and training program. Accept a court-appointed monitor if it uses a third-party pricing algorithm that is not certified pursuant to the terms of the consent decree. Cooperate fully with the states’ ongoing litigation against RealPage and remaining defendants. Colorado will receive $841,500 from the settlement to be held by the attorney general and used for reimbursement of actual costs and attorneys’ fees, future consumer protection or antitrust enforcement, consumer education, or public welfare purposes. This is the third settlement reached in this litigation. Attorney General Weiser announced a settlement with Cortland in April 2025 and a $7 million settlement with corporate landlord Greystar in November 2025. Litigation against RealPage and the remaining property management defendants—Camden, Pinnacle, and Willow Bridge—is ongoing. In securing this settlement, Attorney General Weiser joins the attorneys general of California, Connecticut, Illinois, Massachusetts, Minnesota, North Carolina, Oregon, and Tennessee. ### Media Contact: Lawrence Pacheco Chief Communications Officer (720) 508-6553 office lawrence.pacheco@coag.gov

Matched on: ai_role, home_jurisdiction (AI application: algorithmic_pricing)

#### Attorney General Phil Weiser secures agreement from corporate landlord Cortland to end illegal use of rent pricing software
CO-AG · Colorado · enforcement action · status: unknown · earliest recorded date 11 April 2025

April 11, 2025 (DENVER) – Attorney General Phil Weiser announced today that he has secured an agreement with corporate landlord Cortland Management not to use non-public data from algorithm-driven rent-setting software RealPage. Weiser is joined in the settlement by North Carolina Attorney General Jeff Jackson. Last year, Weiser sued Texas-based revenue management software company RealPage for illegal rental price-fixing that resulted in Coloradans paying millions more in rent. RealPage’s platform relies heavily on landlords who purchase the software sharing private, sensitive property and rental data. After suing RealPage, Weiser later added six large corporate landlords to the suit as defendants, including Cortland. The settlement announced today resolves the allegations against Cortland in the RealPage suit. Previously, Cortland settled with the U.S. Department of Justice over similar allegations. Cortland has also agreed to assist Colorado in its ongoing lawsuit against RealPage. “Coloradans who are struggling to make ends meet are getting hammered by high rent prices, and landlords that collude using private data from RealPage are a part of the problem,” said Weiser. “We are always looking at collaborative solutions when it comes to ensuring a competitive and fair marketplace, and I’m glad that Cortland will no longer be using non-public data from RealPage or software like it to set rents. I will continue to hold accountable any landlord that engages in irresponsible, harmful, and anticompetitive conduct that harms renters by colluding to jack up rents.” Though Cortland will be allowed to use third-party revenue management software to set rents, the types and sources of data they can use will be heavily restricted. As part of the settlement, the company agrees not to use any non-public data from other property management companies to set rents, not to pool or combine non-public data from Cortland-run properties with different owners, and not to share such data with any non-Cortland property owners. Additionally, they will no longer be allowed to use any software that incorporates artificial rent floors or price decrease limits. To ensure cooperation with the settlement, Colorado and North Carolina will have the right to participate in compliance inspections and review compliance information Cortland shares with federal officials under the terms of their settlement with DOJ. Coloradans who believe their landlords are engaging in conduct that violates state law are encouraged to file a complaint with the attorney general at StopFraudColorado.gov Read the consent judgement between the Colorado Attorney General’s Office and Cortland (PDF download). ###

Matched on: ai_role, home_jurisdiction (AI application: algorithmic_pricing)

#### Colorado joins Justice Department in suing six large landlords for algorithmic pricing scheme that harms millions of renters
CO-AG · Colorado · enforcement action · status: unknown · earliest recorded date 7 January 2025

Jan. 7, 2025 (DENVER)—Attorney General Phil Weiser, the Justice Department and a bipartisan coalition of states today filed an amended complaint in its antitrust lawsuit against RealPage (PDF), to sue six of the nation’s largest landlords for participating in algorithmic pricing schemes that harmed renters. The amended complaint alleges the landlords — Greystar Real Estate Partners LLC (Greystar); Blackstone’s LivCor LLC (LivCor); Camden Property Trust (Camden); Cushman & Wakefield Inc and Pinnacle Property Management Services LLC (Cushman); Willow Bridge Property Company LLC (Willow Bridge); and Cortland Management LLC (Cortland) — participated in an unlawful scheme to decrease competition among landlords in apartment pricing, harming millions of American renters. Together, these landlords operate more than 1.3 million units in 44 states and the District of Columbia. Cortland manages over 80,000 rental units in 13 states, including many multifamily apartment buildings in Colorado. Even though the Justice Department also announced that it has reached a settlement with Cortland, Attorney General Weiser said Colorado is not in a position to join it at this time. “Many Coloradans are struggling to afford housing and pay rent. We look forward to learning more about the Cortland settlement and are always willing to consider collaborative solutions. As for the other landlords added to the lawsuit, if they engage in irresponsible and harmful conduct that raises rents, they must be held to account,” said Attorney General Weiser. The amended complaint alleges that the six landlords actively participated in a scheme to set their rents using each other’s competitively sensitive information through common pricing algorithms. Along with using RealPage’s anticompetitive pricing algorithms, these landlords coordinated through a variety of means, including: Directly communicating with competitors’ senior managers about rents, occupancy, and other competitively sensitive topics. Regularly conducting “call arounds” to share, and sometimes discuss, competitively sensitive information about rents, occupancy, pricing strategies, and discounts. Participating in “user groups” hosted by RealPage to discuss how to modify the software’s pricing methodology, as well as their own pricing strategies, including renewal increases, concessions, and acceptance rates of RealPage rent recommendations. In August of 2024, Weiser joined the Justice Department in suing RealPage (opens new tab) alleging that the company enables collusion between landlords and distorts the housing market for millions of Americans who rent. The company does this with software it sells to landlords who agree to share competitively sensitive data, such as rents from executed leases, lease terms, and projected future vacancies. The software combines this massive amount of nonpublic, competitive data collected from competing landlords and feeds it to an algorithm to provide daily, near real-time pricing recommendations back to landlords. Additionally, the software tracks the supply of available apartments, and landlords in markets with heavy demand can use this information to keep properties off the market and drive up rents. The attorneys general of Illinois and Massachusetts joined the amended complaint, along with the original plaintiff states California, Colorado, Connecticut, Minnesota, North Carolina, Oregon, Tennessee, and Washington. ### Media Contact: Lawrence Pacheco Chief Communications Officer (720) 508-6553 office | (720) 245-4689 cell lawrence.pacheco@coag.gov

Matched on: ai_role, home_jurisdiction (AI application: algorithmic_pricing)

#### AG Ferguson investigation shuts down Amazon price-fixing program nationwide
WA-AG · Washington · enforcement action · status: unknown · earliest recorded date 26 January 2022

Amazon will pay $2.25 million, stop its “Sold by Amazon” third-party seller program SEATTLE — Attorney General Bob Ferguson today announced that, as a result of his office’s price-fixing investigation, Amazon will shut down the “Sold by Amazon” program nationwide. The Attorney General’s Office simultaneously filed a lawsuit and a legally binding resolution in King County Superior Court. As part of the legally enforceable consent decree, Amazon must stop the “Sold by Amazon” program nationwide and provide the Attorney General’s Office with annual updates on its compliance with antitrust laws. In addition, Amazon will pay $2.25 million to the Attorney General’s Office, which will be used to support the Attorney General’s antitrust enforcement, which does not receive general fund support. The “Sold by Amazon” program allowed the online retailer to agree on price with third-party sellers, rather than compete with them. Ferguson’s lawsuit asserted that the program violated antitrust laws. Amazon unreasonably restrained competition in order to maximize its own profits off third-party sales. This conduct constituted unlawful price-fixing. Amazon offered the “Sold by Amazon” program from 2018 through 2020 on an invitation-only basis. It invited several hundred third-party sellers with whom it had previously competed for online consumer sales on its online marketplace and other e-commerce platforms. “Consumers lose when corporate giants like Amazon fix prices to increase their profits,” Ferguson said. “Today’s action promotes product innovation and consumer choice, and makes the market more competitive for sellers in Washington state and across the country.” There are about 2.3 million third-party sellers on Amazon worldwide, according to information from a 2018 Amazon letter to its shareholders. Over the last two decades, Amazon’s sales of its own branded products grew from $1.6 billion in 1999 to $117 billion in 2018. Over that same period, third-party sales grew exponentially from $100 million in 1999 to $160 billion in 2018. Third-party sales account for over half the sales on Amazon. Washington state ranks among the top 10 states in the nation with the fastest growing rate of third-party sellers on its online marketplace. Amazon targeted a small fraction of the millions of third-party sellers on its platform to join the “Sold by Amazon” program. Amazon kept the program small as an experiment then slowly began to request more sellers join as it evolved. Ferguson asserted Amazon enticed sellers into the “Sold by Amazon” program by guaranteeing that they would receive at least an agreed upon minimum payment for sales of their consumer goods in exchange for their agreement to stop competing with Amazon for the pricing of their products. Consequently, if sales exceeded the negotiated minimum payment, Amazon and its competitors split the surplus proceeds amongst themselves. For example, if a seller and Amazon agreed to a $20 minimum payment and the item sold for $25, the seller would receive the $20 minimum price and share the $5 additional profit with Amazon, in addition to any fees. The “Sold by Amazon” program resulted in prices for some products increasing when Amazon programmed its pricing algorithm to match the prices that certain external retailers offer to online consumers. As a result, when prices increased, some sellers experienced a marked decline in the sales and resulting profits from products enrolled in the program. Faced with price increases, online customers sometimes opted to buy Amazon’s own branded products — particularly its private label products. This resulted in Amazon maximizing its own profits regardless of whether consumers paid a higher price for sales of products enrolled in the “Sold by Amazon” program or settled for buying the same or similar product offered through Amazon. Prices for the vast majority of the remaining products enrolled in the “Sold by Amazon” program stabilized at artificially high levels. This is because Amazon programmed its pricing algorithm to maintain the seller’s pre-enrollment price as the price floor. This meant participating sellers had limited, if any, ability to lower the price of their products without withdrawing the product’s enrollment in the Sold by Amazon program. For example, while sellers were once able to offer price discounts on their products, Amazon subsequently prevented many sellers from continuing to offer discounts. Sellers then bore the risk of having their products not sell in a timely manner, or at all, while still paying Amazon for things like storage fees of their enrolled products. Many sellers remained stuck with an artificially high price for their products while Amazon was able to maximize its own profits. Assistant Attorneys General Amy Hanson, Rahul Rao, Christina Black, Jonathan Mark, and Eric Newman, economist Ryne Rohla, paralegal Tracy Jacoby and legal assistant Grace Summers are handling the case for the Antitrust Division. Ferguson’s Antitrust Division is responsible for enforcing the antitrust provisions of Washington's Unfair Business Practices-Consumer Protection Act. The division investigates and litigates complaints of anticompetitive conduct and reviews potentially anticompetitive mergers. The division also brings actions in federal court under the federal antitrust laws. It receives no general fund support and funds its own actions through recoveries made in other cases. Amazon has faced allegations of anticompetitive conduct before. In 2013, Amazon abandoned its most-favored nation price parity contract provision across the European Union after British and German government antitrust enforcement officers initiated an investigation into their purpose and effect. Amazon continued using its price parity provision in the US until spring 2019, after scrutiny of it began mounting in 2018. In November 2020, the European Union filed the first-ever antitrust lawsuit against Amazon. Investigators determined Amazon used data it collected from third-party sellers on its marketplace to determine what products to launch and how to price them. Lawsuits have also been filed across the nation since Amazon abandoned its price parity provision in the United States. In January 2021, several consumers filed a price-fixing lawsuit against Amazon in U.S. District Court for the Southern District of New York. The consumers assert Amazon colluded with major publishing firms to drive up the price of e-books. The lawsuit asserts Amazon used anticompetitive contracts to drive up the cost of e-books. In May 2021, District of Columbia Attorney General Karl Racine filed a lawsuit against Amazon for fixing its online retail prices through contract provisions and policies its third-party sellers sign. The provisions and policies, known as “most favored nation” agreements, prevented third-party sellers that offer products on Amazon from offering their products at lower prices or on better terms on any other online platform, including their own websites. In November 2021, Amazon paid $2.5 million for selling highly regulated pesticides on its online platform without a license and without collecting information about their use as required by law. Amazon also sold these regulated pesticides on its site without verifying the licenses of Restricted Use Pesticide purchasers or collecting other legally required information, like the intended use of the pesticide. Amazon failed to inform Washingtonians on the product pages, checkout pages or anywhere else that these regulated agricultural and industrial-use pesticides were different from regular home and garden products. Amazon’s conduct created the impression that anyone could lawfully buy and use the pesticides without restriction. Because of Amazon’s actions, there was no record of how or where the dangerous pesticides were used. As a result of Ferguson’s investigation, Amazon suspended all sales of these pesticides on its site. In May 2019, Ferguson announced the results of an investigation that showed individuals in Washington and across the country made at least 15,188 purchases of products with illegal levels of lead and cadmium from Amazon. Amazon provided more than $200,000 in refunds to consumers and $700,000 to the Attorney General’s Office for future environmental protection efforts. Amazon also entered into a nationwide legally binding agreement to block the sale of children’s school supplies and jewelry on its website without lab reports and other proof from the sellers that the products are not toxic. For information about filing a complaint against Amazon or any other antitrust or consumer protection issue in Washington state, visit https://fortress.wa.gov/atg/formhandler/ago/AntitrustComplaint.aspx. The Office of the Attorney General is the chief legal office for the state of Washington with attorneys and staff in 27 divisions across the state providing legal services to roughly 200 state agencies, boards and commissions. Visit www.atg.wa.gov to learn more. Brionna Aho, Communications Director, (360) 753-2727; Brionna.aho@atg.wa.gov 1125 Washington St SE • PO Box 40100 • Olympia, WA 98504 • 360-753-6200 OFFICE HOURS: 8:00 AM - 5:00 PM Monday - Friday Closed Weekends & State Holidays

Matched on: ai_role, home_jurisdiction (AI application: algorithmic_pricing)

#### Attorney General Jeff Jackson Reaches $7 Million Settlement with North Carolina Landlord to Stop Setting Rent using AI and Non-Public Information
NC-AG · North Carolina · enforcement action · status: unknown · earliest recorded date 18 June 2026

FOR IMMEDIATE RELEASE Thursday, June 18, 2026 Contact: comms@ncdoj.gov 919-538-2809 RALEIGH – Attorney General Jeff Jackson and eight other bipartisan attorneys general today reached a $7 million settlement with LivCor, LLC, a major landlord with approximately 3,500 apartments in North Carolina. LivCor will stop using non-public data from other landlords, either through RealPage’s software or by other means, to set rents. This is the third landlord Attorney General Jackson has reached such a settlement with after suing six property management companies in January 2025 for illegally working together and using RealPage’s AI software to drive up rent for North Carolinians. “One by one, we’re shutting down the illegal scheme that allowed landlords to use illegal AI software to drive up rents for North Carolinians,” said Attorney General Jeff Jackson. “This case is about leveling the playing field so that consumers pay affordable rents.” LivCor will stop: Sharing sensitive data with other landlords. Using third-party software or algorithms to price apartments, unless the software is within the rules. Sharing or using any competitively sensitive data from other landlords and property managers to set rent prices or generate recommended rent prices. Attending or participating in RealPage-hosted meetings of competing landlords. LivCor must report to the Attorney General’s Office on its compliance with the order and allow the Attorney General to participate in inspections to ensure compliance with the judgment’s terms. If necessary, the Attorney General may enforce the agreement in court or seek to extend its terms. Attorney General Jackson’s bipartisan case against RealPage and the remaining three landlords continues. The lawsuit alleges that RealPage used sensitive rental data to build a pricing algorithm that inflated rents and violated antitrust laws. The landlords allegedly shared non-public information with RealPage and each other about rents, occupancy, pricing strategies, and discounts, resulting in higher rents than a competitive market would have produced. Attorney General Jackson has already reached settlements with Greystar and Cortland, the largest and second-largest landlords in the state. The alleged illegal conduct harmed North Carolinians on a wide scale. The seven landlords involved in this case own or manage more than 70,000 rental units across the state. These alleged rent hikes made it harder for North Carolinians to afford their homes at a time when housing costs are already rising, and they also hurt landlords who price their units fairly and follow the law. Attorney General Jackson has been taking the lead on several antitrust fights to protect consumers and fair prices. He recently shut down a secret data exchange between some of the country’s largest meat producers and a company that helped them coordinate to reduce competition and raise prices for chicken, pork, and turkey in North Carolina. He also recently secured a win against Live Nation and Ticketmaster to stop their illegal monopoly over the ticketing industry, which had allowed them to overcharge consumers for years. He also sued to block Nexstar, the nation’s largest television station owner, from buying one of its biggest competitors, Tegna. The lawsuit alleges that the merger would raise prices, weaken local news, and give one company too much control over what communities across the country see on television. Attorney General Jackson is joined in reaching this settlement by the attorneys general of California, Colorado, Connecticut, Illinois, Massachusetts, Minnesota, Oregon, and Tennessee. A copy of the settlement is available here. ###

Relief: Monetary $7 Million

Matched on: ai_role (AI application: algorithmic_pricing)

#### Attorney General Jeff Jackson Reaches $7 Million Settlement with Largest North Carolina Landlord over AI Rent Setting
NC-AG · North Carolina · enforcement action · status: unknown · earliest recorded date 20 November 2025

FOR IMMEDIATE RELEASE Thursday, November 20, 2025 nahmed@ncdoj.gov 919-538-2809 RALEIGH – Attorney General Jeff Jackson and eight other bipartisan attorneys general reached a $7 million settlement with Greystar Management LLC, North Carolina’s largest landlord with more than 25,000 units across the state. Attorney General Jackson sued Greystar and other landlords in January for illegally working together and using RealPage’s AI software to raise North Carolinians’ rents. As part of the settlement, Greystar will stop using non-public data from other landlords, either through RealPage’s software or by other means, to set rents. “This settlement means that more than 25,000 renters in North Carolina are going to be charged fairer prices for rent at a time when housing costs are overwhelming,” said Attorney General Jeff Jackson. “Companies can’t use new technology, like AI, to break the law and hurt customers. If they try, we’ll take them to court.” Greystar will stop: Using sensitive data from its competitors to help set its pricing model. Using third-party software or algorithms to price apartments, unless they do so under the supervision of a court-appointed monitor. Sharing or using any competitively sensitive data from other landlords and property managers to set rent prices or generate recommended rent prices. Attending or participating in RealPage-hosted meeting of competing landlords. Greystar will report to the Attorney General’s Office on how it’s complying with the judgment. The Attorney General will be able to participate in inspections to ensure Greystar is in compliance and, if necessary, can enforce the terms of the agreement in court or extend the term of the agreement. Attorney General Jackson’s bipartisan case against the other four landlords and software company RealPage continues. He is suing RealPage for allegedly exploiting landlords’ competitively sensitive information to create a pricing algorithm that inflated rent prices and violated antitrust laws. Attorney General Jackson’s case alleges that these landlords communicated with RealPage and each other to share non-public information about rent prices, occupancy, strategies for setting rents, and discounts – resulting in higher prices for rent than competitive market forces would have set. Together, these landlords own or manage more than 70,000 units throughout the state. The alleged illegal conduct harms North Carolinians who are struggling to pay rent and stay in their homes as rental prices increase, and they harm landlords who are trying to play fairly and follow the rules. Attorney General Jackson is taking a close look at how to develop safeguards against AI harms to consumers such as rent-setting. Last week, he and Utah Attorney General Derek Brown launched the nationwide bipartisan AI task force to tackle the risks of transformative AI technology. Joining Attorney General Jackson in reaching this settlement were the attorneys general of California, Colorado, Connecticut, Illinois, Massachusetts, Minnesota, Oregon, and Tennessee. A copy of the settlement is available here. ###

Relief: Monetary $7 Million

Matched on: ai_role (AI application: algorithmic_pricing)

#### Texas AG: Google LLC — settlement
TX-AG · Texas · enforcement action · status: settled · earliest recorded date 9 May 2025

Texas AG · settlement · Google LLC · $1,375,000,000 · 2025-05-09

Cited: Texas data privacy laws; Texas CUBI

Relief: Monetary $1,375,000,000

Matched on: home_jurisdiction

Withheld pending data cleanup: forum_name — the matter is reported in full otherwise.

#### Texas AG: Meta (Facebook) — settlement
TX-AG · Texas · enforcement action · status: settled · earliest recorded date 30 July 2024

Texas AG · settlement · Meta (Facebook) · $1,400,000,000 · 2024-07-30

Cited: Texas DTPA; Texas Capture or Use of Biometric Identifier Act (CUBI)

Relief: Monetary $1,400,000,000

Matched on: home_jurisdiction

Withheld pending data cleanup: forum_name — the matter is reported in full otherwise.

#### Justice Department Reaches Proposed Settlement with Willow Bridge, One of America’s Largest Landlords, to Resolve Information Sharing and Algorithmic Coordination Claims
DOJ · federal · enforcement action · status: pending · earliest recorded date 6 July 2026

DOJ · settlement · involving Willow Bridge

Matched on: ai_role (AI application: algorithmic_pricing)

#### Délibération de la formation restreinte n°SAN-2026-008 du 26 mai 2026 concernant la société IQVIA OPERATIONS FRANCE
FR-CNIL · France · enforcement action · status: closed · earliest recorded date 26 May 2026 · docket SAN-2026-008

Matched on: ai_role (AI application: automated_decision)

#### National Consumer Protection Week: Colorado consumers filed record number of complaints in 2025
CO-AG · Colorado · enforcement action · status: unknown · earliest recorded date 2 March 2026

March 2, 2026 (DENVER) – Kicking off National Consumer Protection Week, Attorney General Phil Weiser revealed today that consumers filed nearly 27,000 complaints with his office in 2025. Topping the list are complaints related to retail sales, professional services, real estate, and debt collection scams. In total, Coloradans reported 26,993 complaints with the attorney general’s office in 2025, surpassing the previous record of 24,473 set in 2024. The total number of complaints rose over 10% from 2024. Since 2019, complaints to the Colorado Attorney General’s Office have increased by over 200%. “Every year since I took office, Coloradans have filed a record-setting number of consumer protection complaints. As scams get more advanced with the use of A.I., we want to keep hearing about them,” Weiser said. “Each complaint is valuable to us here at the Colorado Department of Law, as it informs our work and helps us spot trends. Thus far, the department has secured more than $500 million in refunds, restitution, credits, and debt relief for Coloradans during my time as Attorney General.” Every year, the Department of Law recognizes National Consumer Protection Week, an initiative of the Federal Trade Commission, to educate consumers about their rights, how to avoid scams and fraud, and to highlight the importance of filing complaints with law enforcement. The top 10 types of complaints and inquiries received in 2025 are: Top Types of Complaints & Inquiries for 2025 Description 2025 1. Retail Sales These include complaints relating to unauthorized memberships or subscriptions, service and delivery issues, and cancellation/termination issues. 2,064 2. Professional Services These complaints include issues regarding product and service warranties, business support, and legal-related services. 1,631 3. Real Estate Sales & Services These complaints include issues relating to rental and leasing, property management, and real estate related activities. 1,328 4. Debt Collection These complaints include issues under the Colorado Fair Debt Collection Practices Act, including harassment or abuse by a debt collector, disputed debt, and phantom debt or unlicensed collection. 1,227 5. Automobile Sales & Services These complaints include issues relating to automotive purchases, service and repair, auto rentals, and towing services. 1,126 6. Consumer Loans & Credit Sales These complaints include issues under the Uniform Consumer Credit Code, including interest rates, and credit reporting. They also include unlicensed activity, including tribal lending. 915 7. Home Services & Repair These complaints include issues relating to general contracting and remodeling, heating and cooling, and handyman services. 892 8. Imposter Scams In many cases, these scams relate directly to fraudulent telephone calls, emails, or text messages from scammers posing as a government official or employee from a reputable company, often seeking monetary payments through gift cards, wire transfers, and other money transfer services. 868 9. Telecommunications These complaints include issues under the Uniform Consumer Credit Code, including interest rates, and credit reporting. They also include unlicensed activity, including tribal lending. 829 10. Health Care & Medical Services These complaints include issues relating to hospitals and urgent care, health and medical insurance carriers, health practitioners relating to quality of care, medical billing, and coverage issues. 820 Consumer protection actions taken in 2025: Attorney General Phil Weiser, FTC announce $24M settlement with Greystar for misleading renters about costs, saddling tenants with hidden fees AG Weiser secures $400k settlement with Dollar General for overcharging customers Attorney General Phil Weiser stops phony cancer charity from bilking donors (opens new tab) Phony immigration business agrees to shut down operations, pay fines in settlement with Attorney General Phil Weiser Cannabis companies to pay $400,000 fine, cease Colorado operations under settlement with Attorney General Phil Weiser Attorney General Phil Weiser sues to protect Colorado fans from Live Nation and Ticketmaster’s deceptive fees, ticket resale tactics Attorney General Phil Weiser alerts consumers on deleting data shared with 23andMe Bogus blackmail scams on the rise, warns Attorney General Phil Weiser Consumers should report scams, fraud, and other complaints at StopFraudColorado.gov (opens new tab) or by calling 800-222-4444. ### Media Contact: Maxwell Mead Communications Specialist Maxwell.Mead@coag.gov

Matched on: home_jurisdiction

#### AG Brown joins bipartisan group demanding xAI halt creation of nonconsensual sexual content
WA-AG · Washington · enforcement action · status: unknown · earliest recorded date 23 January 2026

Attorney General Nick Brown joined a bipartisan group of 35 attorneys general today demanding that xAI, the company that owns both the X (Twitter) social media platform and the AI chatbot Grok, do more immediately to prevent its chatbot from generating nonconsensual intimate images and child sexual abuse material. In recent weeks, Grok has made this content publicly available at the click of a button, driving harassment and exploitation that deprives people of control over how their bodies and likenesses are portrayed. “We can’t allow big tech companies to flout the law in creating tools that allow users to create nonconsensual intimate images and child sexual abuse material through AI,” said Brown. “I’m heartened by the fact that this bipartisan group can come together to demand accountability for how the platform has facilitated this disturbing behavior.” Users have repeatedly prompted Grok to “undress” individuals, particularly women and children, and to place them in sexualized contexts without consent. In some cases, Grok has generated images depicting children in minimal clothing or sexual situations. The attorneys general note that xAI has marketed Grok’s permissive content generation as a selling point and warn that “the ability to create nonconsensual intimate images appears to be a feature, not a bug.” Although xAI has recently implemented limited measures that appear to have reduced the volume of this content, the attorneys general are demanding assurances that these safeguards are effective, durable, and consistently enforced. They are also urging the company to honor requests to remove this content – a requirement that will soon be mandated under federal law when the Take It Down Act becomes enforceable in May 2026. As the chief law enforcement officers of their states, the attorneys general raise serious concerns that Grok’s outputs may violate state and federal civil and criminal laws governing nonconsensual intimate images, the creation and distribution of child sexual abuse material, and the legal remedies available to victims. In Washington state, disclosing intimate images of another person without their consent is illegal, including fabricated images, sometimes referred to as “deepfakes.” Possession or distribution of child sexual abuse material is also a crime under Washington state law, which also includes fabricated depictions. The attorneys general are demanding that xAI share how it intends to: Brown joins the attorneys general of North Carolina, Utah, Pennsylvania, Connecticut, American Samoa, Arizona, Colorado, Delaware, District of Columbia, Hawaii, Idaho, Illinois, Kansas, Kentucky, Maine, Maryland, Michigan, Minnesota, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Dakota, Northern Mariana Islands, Oklahoma, Oregon, Rhode Island, South Dakota, Vermont, Virgin Islands, Virginia, Wisconsin, and Wyoming in sending this letter. Washington’s Attorney General serves the people and the state of Washington. As the state’s largest law firm, the Attorney General’s Office provides legal representation to every state agency, board, and commission in Washington. Additionally, the Office serves the people directly by enforcing consumer protection, civil rights, and environmental protection laws. The Office also prosecutes elder abuse, Medicaid fraud, and handles sexually violent predator cases in 38 of Washington’s 39 counties. 1125 Washington St SE • PO Box 40100 • Olympia, WA 98504 • 360-753-6200 OFFICE HOURS: 8:00 AM - 5:00 PM Monday - Friday Closed Weekends & State Holidays

Matched on: home_jurisdiction

#### Datenschutzbehörde 2026-0.018.391
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 9 January 2026 · docket 2026-0.018.391

Matched on: ai_role (AI application: automated_decision)

#### Délibération de la formation restreinte n° SAN-2025-017 du 30 décembre 2025 concernant la société X
FR-CNIL · France · enforcement action · status: closed · earliest recorded date 30 December 2025 · docket SAN-2025-017

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.395.497
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 15 December 2025 · docket 2025-0.395.497

Matched on: ai_role (AI application: automated_decision)

#### Délibération de la formation restreinte n° SAN – 2025-014 du 11 décembre 2025 concernant la société MOBIUS SOLUTIONS LTD
FR-CNIL · France · enforcement action · status: closed · earliest recorded date 11 December 2025 · docket SAN-2025-014

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.317.107
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 14 November 2025 · docket 2025-0.317.107

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.923.556
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 14 November 2025 · docket 2025-0.923.556

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.914.457
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 10 November 2025 · docket 2025-0.914.457

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.682.914
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 3 November 2025 · docket 2025-0.682.914

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.119.280
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 27 October 2025 · docket 2025-0.119.280

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.818.263
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 15 October 2025 · docket 2025-0.818.263

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.328.963
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 14 October 2025 · docket 2025-0.328.963

Matched on: ai_role (AI application: automated_decision)

#### Datenschutzbehörde 2025-0.355.103
AT-DSB · Austria · enforcement action · status: closed · earliest recorded date 25 September 2025 · docket 2025-0.355.103

Matched on: ai_role (AI application: automated_decision)

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: Real matters with source URLs, de-duplicated to one row per source document.

Limits of this finding:

• 3 entr(y/ies) have one or more display fields withheld pending cleanup; the matter itself is unaffected.

5. Obligation areas with no matched matter on record

Read this as a statement about the record, in one direction only: these are areas this profile sits in where the enforcement corpus this product holds contains no matched matter. It is neither an assurance nor a warning — an area can be empty because little has been brought, or because our coverage of it is thin.

Matched matters on record (2)

• Pricing conduct and algorithmic coordination — 15 matched matters, listed in section 4.
• Automated decision-making disclosure — 65 matched matters, listed in section 4.

What this was measured against

Corpus held on the run date: 246 matters, of which 96 matched this profile and 90 of those carry a provable date. Computed against the enforcement corpus this product holds, on the run date. It is not a statement about US enforcement in general, and absence here is absence of a record, not absence of risk.

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: Set difference between the obligation areas this company is in and the areas with a matched matter in the corpus we hold.

6. This company in the AI enforcement corpus

3 matters in the AI enforcement corpus carry this name. A name match is not an identification: corporate families share names and a caption can name a company that is not a party. Each record links to its source for confirmation.

A name match is not an identification. Corporate families share names and a caption can name a company that is not a party — each row links to its source so the reader can confirm.

Confidence: LOW — Low — thin evidence, usable as a pointer only

Basis: Tiered name match against published organisation parties and matter captions in the AI enforcement corpus.

Limits of this finding:

• 2 match(es) rest on a caption mention with no published party row, which is the weakest tier and may name a company that is not a party.

7. How much of AI enforcement sits in this area

96 of 246 matters (39%) in the AI enforcement corpus match this profile.

Within those, Austria data protection authority accounts for 37 of 96 matched matters (38.5%); the most-represented AI application among them is automated_decision (65 of 96).

All-time, with no time dimension. It shows how the matched matters are distributed, not whether activity is increasing. The flow question belongs to finding 3, which has its own reliability gate.

Computed over the 246 canonical matters this product holds on the run date. A share here reflects both what regulators have done and what this corpus covers; the two cannot be separated from inside the data, and neither is reported as the other.

Confidence: HIGH — High — multiple records, primary sources

Basis: Share of matched matters over the canonical AI enforcement corpus. Stock, not flow.

8. What moved in these instruments

660 substantive revisions in the jurisdictions on this profile. A further 0 revisions touched only our own metadata (source URL, cached payload) and are excluded as records of re-reading rather than of legal change.

Substantive revisions (30 of 660 shown)

observed_at is when THIS PRODUCT saw the revision, not when the change was made. It is reported as such and no rate or trend is derived from it. Ordering uses the instrument's own record_date where one exists.

The revision log was read up to 4000 rows, most-recently-observed first, and that ceiling was reached. Older revisions exist and were not examined, so this is the recent tail of the log rather than its whole history.

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: Substantive-field revisions only (status, lifecycle, in-force, dates, text). Metadata churn excluded.

Limits of this finding:

• The revision log was read up to 4000 rows, most-recently-observed first, and that ceiling was reached. Older revisions exist and were not examined, so this is the recent tail of the log rather than its whole history.

9. Freshness of the data behind this brief

38 sources cover the jurisdictions on this profile. 36 are reading normally; 2 are past their own silence threshold and 2 have never returned a successful read. Findings drawn from a stale or never-read source are limited by that, and the affected sources are named below rather than averaged into a single freshness figure.

Sources not reading normally

Sources reading normally that carry nothing for this profile

These feeds are being read successfully and have produced no matter matching this profile. That is a measured absence, not a gap in collection — the distinction matters, because an empty section states neither.

In scope: 38 sources. Healthy 36 · stale 2 · never succeeded 2 · reading but carrying nothing for this profile 4. A further 224 sources cover jurisdictions not on this profile and 119 are disabled; neither group is shown.

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: Per-source ingest health, using the silence threshold each source declares for itself in the database.

Limits of this finding:

• 2 in-scope source(s) are past their own silence threshold, and findings drawn from them are limited by that.
• 4 in-scope source(s) are reading normally and carry no matter matching this profile — a measured absence rather than a collection gap, listed by name.
• 2 in-scope source(s) have never returned a successful read, so the jurisdictions they cover may be under-represented throughout this brief.

10. This sector against the others

Real Estate carries 0 of 246 canonical matters (0%), placing it 10 of 15 mapped sectors. Matters count toward every sector they touch, so these shares overlap.

Confidence: NONE — Not established — see the limits below

Basis: Share of the canonical enforcement corpus attributable to each sector, by law-domain branch.

Limits of this finding:

• no records under this finding
• A matter counts toward every sector it touches, so these shares overlap and do not sum to 100%.

11. Where the exposure sits

Federal instruments reach every state on this profile and are counted once here rather than repeated per state. Federal instruments reaching this company: 4.

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: State-layer instruments from finding 1, matched matters by the matter's own jurisdiction. Federal instruments counted once, not per state.

Limits of this finding:

• 2 declared state(s) carry no binding AI instrument in the corpus — an absence of statute, not an absence of obligation, since the federal layer still reaches them.

12. Where federal and state authority overlap

4 federal and 3 state instrument(s) reach this company at the same time. That is an overlap of authority, not a finding of conflict: whether two instruments conflict, and which would prevail, is a legal conclusion a court reaches and is not something this data can establish.

Federal instruments

• Equal Credit Opportunity Act (ECOA)
• Fair Credit Reporting Act (FCRA) § 1681
• Fair Housing Act (FHA)
• HUD AI Guidance (2023)

State instruments

• SB 942 — AI Transparency Act — California
• TRAIGA — Texas Responsible AI Governance Act (HB 149, 2025) — Texas
• SB 24-205 — Colorado AI Act (amended 2026 by SB 26-189) — Colorado

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: Instrument layer within the applicable set. No per-application mapping is invented.

Limits of this finding:

• Overlap of authority is established; whether the instruments CONFLICT, and which would prevail, is a legal conclusion this data cannot reach.

13. What is moving

1 dated commencement(s) and 439 bill(s) whose status moved. Every row is something that already exists and already carries the date or status shown — none of it is a forecast.

Nothing in this table is a forecast. Each row reports a date or a status already on the record; a bill that is moving may never become law.

Confidence: MEDIUM — Medium — real records with a stated limitation

Basis: Future-dated commencements from finding 2 plus pending bills whose status moved. Nothing here forecasts an outcome.

Limits of this finding:

• 439 row(s) are bills that have NOT become law and may never; only the movement is a fact.

Method and limits

No AI guesswork. Every finding is computed deterministically from primary sources and linked to them. No language model interprets the law.

This brief is generated by rule from records held in the AI Law Tracker corpora. deterministic rules over retrieved records; no model in this path. Findings are reproducible: the same profile against the same data produces the same brief.

Methodology

Findings are computed, not generated. Every finding is derived deterministically from primary sources and linked to them. No language model is called at any point in producing this brief — it does not write the findings, and it does not phrase them either.

Where the data comes from. Two corpora, both maintained by us and both built from primary sources. The AI-law corpus holds statutes, regulations and bills collected from official legislature and agency sites. The enforcement corpus holds actions brought by regulators — federal agencies, state attorneys general, and named foreign authorities — each recorded with the agency, the instrument, the AI application, and a link to the announcement it came from. Nothing is entered from a secondary summary without being marked as such.

How a company is matched to the record. A profile carries a sector, the states it operates in, and the AI applications it runs. Instruments are resolved against it by rule — jurisdiction, layer, and whether the instrument is in force on a proven date. Enforcement matters are matched on four axes: AI application, the statute cited, the home jurisdiction, and sector. The first three are treated as strong signals and sector as a weak one, because a sector match alone describes an industry rather than a company.

How the instrument text is read. When a brief is produced, the official source of each applicable instrument is read at that moment and is not stored. Provisions that impose an obligation and name the subject matter of the profile are quoted word for word, with the citation and a link to the document the words came from. Whitespace is normalised; nothing else is changed, and nothing is paraphrased.

When the engine declines to answer. A direction of travel for enforcement ("rising", "falling") is stated only when four conditions hold at once: enough dated matters, enough of them dated, enough distinct months, and comparable coverage across the two windows. That last condition exists because a collection lane going quiet looks exactly like enforcement declining. A monetary range is given only over matters that actually record an amount, and the count is printed beside it. A penalty recorded in a source that covers several matters is not attributed to any one of them. A provision whose citation is not unique in its document is dropped rather than guessed. Where a source could not be read, the brief says so instead of reporting an empty result.

What is not in the pipeline. No language model. Not to classify a matter, not to summarise a statute, not to phrase a finding. Every sentence in a brief is assembled from values the engine computed, by templates held in the codebase, and a test fails the build if a model call appears anywhere in the intelligence engine. This also means the brief contains no assessment of whether an obligation has been met, no prediction of enforcement, and no scoring of any individual person.

Scope

• Corporate only. Nothing in this engine scores, profiles or evaluates an individual person.
• Company and case questions are answered from the AI enforcement corpus, which is scoped to AI matters and is not a general litigation search.
• Coverage is the United States plus the foreign jurisdictions named in the profile, limited to the AI-law corpus this product maintains.
• Enforcement findings are computed against the matters this product holds on the run date, which is not the universe of US enforcement.

Enforcement corpus state at run time

• Read: live
• Matters held: 246; matched to this profile: 96; duplicate rows excluded: 0
• Match axes available: ai_role, statute, home_jurisdiction, sector
• The statute axis matches the verbatim citation_text on each matter, not its linked corpus record: matter_laws links resolve only to USC title level, which is too coarse to establish that two parties were charged under the same provision.

Data sources available to this run

• AI enforcement matters: available — alt-enforcement · v_public_matters and its export views
• Curated AI statute corpus: available — src/data/laws.js and src/data/federalContext.js
• Instrument change history: available — main ALT · legal_records_history
• Source health ledger: available — alt-enforcement · v_source_health
• US case-law party lookup: available. AI-vertical opinion corpus: 37 opinions confirmed against their own full text, 431 recorded citations, and 7 further row(s) held back for review because their text carries no AI term. Identified from CourtListener cluster metadata, so it covers cases whose metadata names the subject — it is not a complete index of every US opinion touching AI.

What this product does not do

• No risk score, rating, ranking or profile of an identified or identifiable natural person.
• No score, tier or rate factor intended to price or select an insurance risk.
• No sale, export or onward supply of personal data, and no building of a person-keyed dataset.
• No unsolicited approach to a LAWYER or firm offering to sell what the engine found in a specific matter they are acting in. Telling a COMPANY what the public record shows about its own sector, its own states, and its own matters is ordinary marketing and is allowed.

Automated analysis of public legal data. This states what the data shows; it is not legal advice, not an assessment of whether any obligation has been met, and not a prediction of enforcement. Every finding links to its source. Decisions should be taken with qualified counsel.

Regulatory Exposure Intelligence reports what the public legal record shows. It is not legal advice, not an assessment of whether any obligation has been met, and not a prediction of enforcement. Every finding links to its source. Decisions should be taken with qualified counsel.