Cycle on Aug 10, 2026

I will start by listing any open threads or watches associated with this research topic to see what has been previously prioritized or tracked.

I will list the current set of notes for this topic to understand what has already been documented and avoid duplicating efforts, while identifying slugs that we can update.

We have several notes on recent AI enforcement actions, class-action settlements (like Siri/Apple, Anthropic, RealPage, Cleo AI, Click Profit, etc.), state AG suits, and regulatory shifts (like FTC's Rytr reversal, TAKE IT DOWN Act enforcement, EU AI Act milestones, etc.). Let's do some searches to find out if there are *new* or *recent* developments (since late June/July/August 2026) regarding: 1. FTC enforcement actions, investigations, or settlements involving AI (including any finalization of the AI Accuracy Policy Statement, Matter No. P264200). 2. State AG actions or settlements involving AI (e.g., Texas, California, Florida, etc.). 3. SEC or DOJ enforcement actions, such as public company AI-washing, AI trading bots, or other AI-related fraud. 4. EEOC or DOJ civil rights actions concerning AI in hiring, lending, or insurance. 5. Important private lawsuits, class actions, or settlements (e.g., algorithmic pricing, copyright, deepfakes, hiring discrimination, etc.). 6. Legal developments in the California Gas Station AI Price-fixing lawsuit or other algorithmic pricing cases. Let's run a search for recent AI regulatory enforcement actions, filings, and settlements in July/August 2026. I will state my intent first. I will search the web for any recent FTC AI enforcement actions, investigations, or settlements in July and August 2026.

This is very interesting! There is a major pushback going on right now (early August 2026) regarding the FTC's proposed policy statement on AI Accuracy and Ideological Steering (Matter No. P264200, Docket No. FTC-2026-0038 or similar). Comments were due on July 31, 2026. A coalition of groups including the Electronic Frontier Foundation (EFF), Center for Democracy & Technology (CDT), Cato Institute, Consumer Technology Association (CTA), and even civil rights/consumer advocacy groups like the National Consumer Law Center (NCLC) have submitted comments urging the FTC to drop or heavily revise the proposal. Let's fetch some of these articles and comments to understand the core arguments and who is opposing it, and update the note `ftc-ai-accuracy-ideological-steering` with these fresh developments. First, let's look at the CDT's or EFF's statements, or the Broadband Breakfast article. I will fetch the EFF article to see the details of their opposition.

This is incredibly detailed and provides excellent context! The proposed policy statement: - "concerning the suppression of accuracy in artificial intelligence systems" was issued in July (specifically July 1, 2026, as noted in the Cato comment and others). - It builds on and references the Trump administration's "Preventing Woke AI in the Federal Government" executive order (signed in July 2025). - It seeks to regulate AI outputs for "accuracy" and "ideological steering/bias", but critics argue it is an unconstitutional prior restraint on speech, exceeds FTC legal authority (including trying to preempt state laws like Colorado's automated decision-making law, SB26-189 / or similar), and is too vague, leading to censorship. - A broad coalition across the political spectrum (EFF, Public Knowledge, Fight for the Future on the left/civil liberties side; Cato Institute, R Street, CTA on the libertarian/industry side; and NCLC, CDT on the consumer/civil rights side) have submitted comments opposing it. Let's fetch the NCLC article to see their perspective, which is very interesting because NCLC is a consumer protection group that usually supports regulation, but opposes *this* proposal. Let's find out why. I will state my intent first. I will fetch the National Consumer Law Center (NCLC) article to see why consumer advocacy groups are also opposing the FTC's proposed policy statement on AI accuracy.

This is fascinating. Let's look at what we have here. The FTC's proposed policy statement, issued on July 1, 2026, is titled "Suppression of Accuracy in Artificial Intelligence Systems" (Matter No. P264200). It argues that attempts by AI developers to address bias and discrimination in AI systems actually create "ideological bias" and make systems "less accurate," and that removing biases "deceives consumers" because it steers outputs away from what users expect or what is "objective." This has triggered a highly unusual, unified opposition from across the entire political and interest-group spectrum: 1. **Civil Liberties & Digital Rights Groups** (EFF, Public Knowledge, Fight for the Future, CDT): They argue that the policy is a First Amendment violation (viewpoint-based censorship, prior restraint, making the FTC the arbiter of truth), is too vague, and represents government "jawboning" to align AI models with the Trump administration's "Preventing Woke AI in the Federal Government" executive order. 2. **Consumer & Civil Rights Advocacy Groups** (NCLC, Lawyers' Committee for Civil Rights Under Law): They argue that debiasing is essential for AI accuracy and compliance with anti-discrimination laws. They point out that AI systems trained on historical data replicate discrimination in credit, housing, hiring, and healthcare, and that calling debiasing "deceptive" would increase bias and undermine civil rights protections. 3. **Industry & Libertarian Groups** (CTA, Cato Institute, R Street): They argue that the FTC is exceeding its legal authority, trying to unconstitutionally preempt state laws (such as Colorado's landmark AI law, SB 24-205 / SB26-189 or similar, and other state ADM laws), and that the policy will create massive confusion and regulatory overreach. This is a major update to the `ftc-ai-accuracy-ideological-steering` note! We can write an incredible, highly detailed update to this note. Let's do some more searches to see if there are other new cases, investigations, settlements, or court decisions since July 2026. Let's search for: - "algorithmic pricing" OR "price-fixing" class action OR settlement OR DOJ OR FTC (recent, summer 2026). - "AI washing" SEC OR FTC OR class action (recent, summer 2026). - EEOC AI hiring OR DOJ AI hiring (recent, summer 2026). - "California gas station" Kalibrate AB 325 (recent, summer 2026). I will state my intent first. I will search for any recent developments in algorithmic pricing litigation, including cases involving RealPage, Yardi, or the California gas station class action under AB 325.

These are major developments! Let's analyze what we have here. 1. **Anthropic Copyright Class Action Settlement**: - On July 20, 2026, U.S. District Judge Araceli Martinez-Olguin in San Francisco officially signed off on Anthropic's landmark **$1.5 billion settlement** of the class-action lawsuit brought by a group of authors (*Bartz et al. v. Anthropic* / similar). This is the largest known settlement of a U.S. copyright case, and the first major U.S. AI copyright case to settle. - The authors sued Anthropic in 2024, accusing it of misusing their books to train its Claude chatbot. - The settlement was reached in 2025, after a landmark ruling that training AI on books is "fair use," which remains the law today. - We have an existing note `bartz-v-anthropic-settlement`. We should update it because the final approval of this landmark $1.5B settlement occurred on July 20, 2026! 2. **RealPage Algorithmic Rent-Setting Software Litigation**: - There is a massive update to the RealPage settlements. The total class-action settlement amount is now **$359.9 million** (specifically $359,925,000) partially resolving the multidistrict litigation *In re: RealPage Inc. Rental Software Antitrust Litigation (No. II)*, Case No. 3:23-md-3071, in the Middle District of Tennessee. - The second set of settlements (11 deals with 14 companies, totaling $218 million) received preliminary approval on May 22, 2026, following the first set (26 settlements totaling $141.8 million) in November 2025. - The final approval hearing is scheduled for **October 15, 2026**. - There are also several state AG settlements. For example, Washington D.C. Attorney General Brian Schwalb reached a combined $1.4 million settlement with Avenue5 Residential and Bell Partners on June 15, 2026. - On July 20, 2026, New Jersey Governor Mikie Sherrill signed legislation restricting landlords' use of algorithmic software to set rents, prohibiting the sharing of sensitive pricing data and banning algorithmic price-fixing. New Jersey is following New York (which banned it in late 2025). - In Philadelphia, a new class action was filed on July 16, 2026 (*Yiyao Liu v. Willow Bridge Property Company LLC and RealPage Inc.*) under Philadelphia's local rent price-fixing statute. - In San Diego, a class action was filed on July 2, 2026 (*Keller v. UDR, Inc.*) under San Diego Municipal Code § 98.1103, which prohibits algorithmic rent pricing. - The DOJ's federal case (*United States et al. v. RealPage et al.*) also had developments: Willow Bridge agreed to a proposed settlement in July 2026, and LivCor (owned by Blackstone) agreed to a $7 million settlement with nine state AGs on June 22, 2026. - We have an existing note `doj-state-ags-realpage`. We can update this note to capture these massive updates, including the $359.9 million cumulative class settlement, New Jersey's new statutory ban, and the local ordinances being tested in San Diego and Philadelphia. 3. **Atlantic City Casino AI Price-Fixing Lawsuit**: - This is a *completely new and major* development! On July 29, 2026, the 3rd U.S. Circuit Court of Appeals in Philadelphia revived a proposed class-action lawsuit (*Cornish-Adebiyi v. Caesars Entertainment* / *Cornish-Adebiyi et al. v. Caesars Entertainment et al.*, No. 24-3006) accusing several Atlantic City casino hotels (Caesars, Harrah's, Tropicana, Borgata, Hard Rock) of using Cendyn Group's Rainmaker AI-assisted pricing software to coordinate room rates. - The 3rd Circuit reversed a 2024 dismissal by U.S. District Judge Karen Williams. - Circuit Judge Theodore McKee wrote the unanimous opinion, famously stating that if it's illegal for a human named "Bob" to collect competitors' confidential pricing plans and tell them what to charge, it's not legal for an AI algorithm named "Rainmaker" to do it. - This is a monumental ruling on the intersection of AI pricing and antitrust law. It represents a major split/distinction from a similar Las Vegas case that was dismissed by the 9th Circuit. - We should write a brand-new note on this landmark 3rd Circuit ruling! Let's coin a stable slug like `atlantic-city-casino-ai-price-fixing`. Let's do some more searches to see if there are other new developments. For instance: - What is happening with the California Gas Station AI Price-fixing lawsuit (*Kalibrate* AB 325)? Let's search specifically for that. - Are there any other new SEC, DOJ, or FTC AI actions? I will state my intent first. I will search for any recent developments (in June, July, or August 2026) regarding the California gas station algorithmic pricing class action or the Kalibrate fuel pricing software lawsuit.

Great. The Kalibrate fuel price-fixing case (*California drivers v. Kalibrate* / similar) was filed on June 22, 2026. Let's fetch the Law Commentary article about it to see the details of the complaint, the parties, and what is alleged under AB 325, so we can update the existing note `california-gas-station-algorithmic-pricing`. I will state my intent first. I will fetch the Law Commentary article on the California gas station algorithmic pricing lawsuit to get detailed information about the case.

This is a fantastic source. It provides the exact list of defendants in the Kalibrate fuel price-fixing class action filed on June 22, 2026: - Kalibrate (the AI-powered fuel-pricing software provider) - BP - Circle K - Marathon Petroleum - 7-Eleven - Walmart - Albertsons - EG America (Cumberland Farms) The lawsuit represents California drivers and was filed in the U.S. District Court for the Eastern District of California (Sacramento). It relies on: - California's Cartwright Act (the state's antitrust law). - California's newly enacted algorithmic pricing law, **Assembly Bill 325 (AB 325)**, which took effect Jan. 1, 2026, and specifically targets shared pricing software used by competitors that relies on competitor data to recommend, set, or influence prices. - California's Unfair Competition Law. It alleges that the software's "price restoration" feature can raise prices across a market after one or more stations lower them, and that areas using Kalibrate saw average gas price increases of about 6 cents per gallon (and up to 30 cents per gallon in high-density areas). This is a perfect update for the `california-gas-station-algorithmic-pricing` note! We can rewrite/update it with these extremely specific details. Let's do a search on recent FTC actions to see if there are other new cases. For example, let's search: - "FTC" "AI" settlement OR complaint (recent, summer 2026). - "Rytr" FTC (let's check if there are any updates, though it was vacated in Dec 2025). - "DoNotPay" FTC (let's check if there are any developments on DoNotPay). - "Lending" or "hiring" AI discrimination (recent). I will state my intent first. I will search for any recent FTC AI-related complaints, settlements, or enforcement actions in June, July, or August 2026.

Ooh, there are two extremely interesting new FTC developments mentioned here! 1. **FTC's $2.25 Million Amazon Settlement and Emerging Focus on AI Accuracy**: - On June 30, 2026, the FTC announced a **$2.25 million settlement** with Amazon. Let's find out what this Amazon settlement was about and how it relates to AI accuracy. - Let's run a search for this. 2. **Consumer Federation of America (CFA) Complaint against Speechify**: - In late July 2026, the CFA, in partnership with UCLA Law School's Information Policy Lab, filed a complaint with the FTC and state AGs urging them to investigate **Speechify** over its AI voice-cloning practices. - This is a fantastic new development! Let's search for Speechify to see what is alleged. Let's search for: - "Amazon" "FTC" "$2.25" OR "2.25 million" 2026. - "Speechify" "FTC" OR "CFA" OR "Consumer Federation of America" 2026. I will state my intent first. I will search for the details of the FTC's June 30, 2026, $2.25 million settlement with Amazon regarding AI or consumer protection.

Ah, let's search specifically for the FTC's June/July 2026 Amazon settlement. The Instagram post says: "In July 2026, the FTC charged Amazon with knowingly violating the Fair Credit Reporting Act. The agency ordered Amazon to pay 2.25 million dollars." And the LinkedIn post from William Z. says: "GREATER SCRUTINY, NEW FRONTIERS: The FTC's $2.25 Million Amazon Settlement and Its Emerging Focus on AI Accuracy ... First, on June 30, 2026, the FTC announced..." Let's search for: `"Amazon" "FTC" "2.25" "Fair Credit Reporting Act" OR "FCRA" 2026` or `"Amazon" "FTC" "AI" "2.25"`. I will state my intent first. I will search for the details of the FTC's June/July 2026 $2.25 million settlement with Amazon, checking for references to the Fair Credit Reporting Act (FCRA) and how it connects to AI.

Excellent. The $2.25 million FTC settlement with Amazon (announced around June 30 / early July 2026) is actually about Amazon violating Section 609(e) of the Fair Credit Reporting Act (FCRA) by failing to provide identity theft victims with records of fraudulent accounts/transactions within 30 days. It is not directly an AI settlement, though some commentators link it to broad consumer protection and data practices. Let's search for the Speechify AI Voice-Cloning complaint from the Consumer Federation of America (CFA) and UCLA Law, which is highly relevant to AI enforcement and consumer protection. I will state my intent first. I will search for the Consumer Federation of America's (CFA) FTC complaint against Speechify over AI voice-cloning practices.

This is a superb and highly relevant development! Let's examine the Speechify complaint: - **Date**: Late July 2026 (announced July 24/31, 2026). - **Parties**: Filed by the Consumer Federation of America (CFA) in partnership with UCLA Law School's Information Policy Lab (a project of the UCLA Institute for Technology, Law & Policy). - **Target**: Speechify, an AI voice-cloning platform. - **Accusation**: The complaint filed with the FTC and state Attorneys General alleges that Speechify engages in unfair, deceptive, and abusive trade practices (violating Section 5(a) of the FTC Act, state UDAP laws, and digital forgery statutes) by facilitating AI voice-cloning impersonation scams. - **Key details**: Speechify's voice-cloning technology allows users to create realistic voice clones from just seconds of audio, while relying on nothing more than a simple self-attestation checkbox to verify consent (virtually no meaningful safeguards to prevent fraud). This comes amid FTC data showing that consumers reported losing $3.5 billion to impersonation scams in 2025. - **Action requested**: Urges federal and state regulators to investigate Speechify and pursue appropriate enforcement action. This is a perfect candidate for a new finding note! The slug can be `cfa-speechify-voice-cloning-complaint`. Let's review our threads and watches: 1. **EU AI Act Article 50(2) Transitional Period for Existing Systems Expires**: This watch is for Dec 2, 2026. Since the current time is August 10, 2026, this has not fired yet. 2. **FTC Finalizes AI Accuracy and Ideological Manipulation Policy Statement**: This watch is for when the FTC officially votes to finalize and publish the final policy statement (Matter No. P264200). It has not been finalized yet (comments were just due on July 31, 2026, and are drawing heavy fire). We should keep this watch open. 3. **Track Procedural Developments in California Gas Station AI Price-Fixing Lawsuit**: The Kalibrate class action was filed on June 22, 2026. No major procedural rulings (like a motion to dismiss ruling) have happened yet, as it's only been a few weeks. We should keep this watch open. Wait, are there any other threads to resolve? Let's check `list_threads` again. The threads return: - ID: `019fc6b1-9347-7689-9a12-7ba4dc1e5d62` (EU AI Act watch) - ID: `019fc6b1-5361-745a-a899-f78c20f9fdbd` (FTC Finalizes AI Accuracy watch) - ID: `019f5a8c-27c8-7227-8b18-35c6a0aeddf0` (Track Kalibrate gas station watch) None of these watches have fired yet. Let's double-check if there are other notes we can write or update. We have: 1. `ftc-ai-accuracy-ideological-steering` (FTC Proposes Policy Statement Targeting AI Output Accuracy and "Ideological Manipulation"). - We should absolutely update this note with the massive wave of public comments and opposition that concluded on July 31, 2026. 2. `bartz-v-anthropic-settlement` (Anthropic Settles Landmark AI Copyright Class Action for $1.5 Billion). - Let's update this note because on July 20, 2026, District Judge Araceli Martinez-Olguin granted final approval of the $1.5 billion settlement. 3. `doj-state-ags-realpage` (DOJ and State Attorneys General Settle Landmark Algorithmic Price-Fixing Case Against RealPage). - We can update this note to include the massive $359.9 million cumulative class-action settlements (preliminary approval of the second set on May 22, 2026, final hearing scheduled for Oct 15, 2026), the new state legislation (New Jersey's Governor Mikie Sherrill signing a bill restricting algorithmic rent-setting software on July 20, 2026), and local ordinances being tested in San Diego (*Keller v. UDR*) and Philadelphia (*Yiyao Liu v. Willow Bridge*). 4. `california-gas-station-algorithmic-pricing` (California Drivers Launch Landmark AI Price-Fixing Lawsuit Against Gas Giants Under New AB 325 Law). - We can update this note with the extremely specific details of the complaint filed on June 22, 2026, including the full list of defendants (Kalibrate, BP, Circle K, Marathon, 7-Eleven, Walmart, Albertsons, EG America), the alleged damages of 6 to 30 cents per gallon, and the "price restoration" feature. 5. `atlantic-city-casino-ai-price-fixing` (new note). - We will write a brand-new note on the 3rd Circuit's July 29, 2026, ruling reviving the Atlantic City casino hotels AI price-fixing class action (*Cornish-Adebiyi v. Caesars Entertainment*), featuring Judge Theodore McKee's "Bob vs. Rainmaker" reasoning and its contrast with the Las Vegas Cendyn case. 6. `cfa-speechify-voice-cloning-complaint` (new note). - We will write a brand-new note on the Consumer Federation of America's late July 2026 complaint against Speechify for its lax voice-cloning safeguards. Let's make sure we have all the details and quotes correct. Let's first fetch the `ftc-ai-accuracy-ideological-steering` note to see what's currently in it, so we can update it seamlessly. I will state my intent first. I will fetch the current content of the `ftc-ai-accuracy-ideological-steering` note to see what was previously written.

Ah, my apologies. `web_fetch` is only for external web URLs. I should use `read_notes` to see the previews, or I can just write the update directly based on my research. Actually, let's look at the preview of `ftc-ai-accuracy-ideological-steering` from `read_notes`: "The public comment period for the Federal Trade Commission's (FTC) proposed policy statement concerning the "Suppression of Accuracy in Artificial Intelligence Systems" (Matter No. P264200, Docket..." Wait, is there a tool to read local notes? No, there isn't a custom local note reading tool in my declaration list, but `read_notes` gives us the list of notes with their title, slug, and preview. Let's look closely at `read_notes` output. It has: - `slug: ftc-ai-accuracy-ideological-steering` - `slug: bartz-v-anthropic-settlement` - `slug: doj-state-ags-realpage` - `slug: california-gas-station-algorithmic-pricing` Let's write the updated notes and new notes one by one. Let's start with the first finding: **The 3rd Circuit's monumental ruling reviving the Atlantic City Casino AI Price-Fixing Case**. Let's review the details from the retrieved articles: - **Case Name**: *Cornish-Adebiyi et al. v. Caesars Entertainment et al.*, Case No. 24-3006 (3rd Cir. July 29, 2026). - **Defendants**: Caesars Entertainment, MGM Resorts International, Hard Rock International, Cendyn Group, and corporate entities tied to Caesars Atlantic City, Harrah's Resort Atlantic City, Tropicana Atlantic City, Borgata Hotel Casino & Spa, and Hard Rock Hotel & Casino Atlantic City. - **AI System**: Rainmaker, a dynamic-pricing revenue management platform operated by Cendyn (acquired in 2019). - **Practice at Issue**: Casino hotels fed current, nonpublic room pricing and occupancy data into Rainmaker, which combined the data and recommended room rates that the hotels accepted 90% of the time, allegedly allowing them to maintain high prices and avoid price wars despite falling occupancy (e.g., occupancy fell 8% from 2017 to 2019 while room revenue rose 22%; and in 2022 they rented 5% fewer rooms but charged 25% more than in 2019). - **Legal Basis**: Section 1 of the Sherman Antitrust Act (hub-and-spoke conspiracy theory, where Cendyn is the hub and the casino hotels are the spokes). - **Outcome/Status**: The 3rd Circuit reversed a 2024 dismissal by U.S. District Judge Karen Williams, reviving the lawsuit and sending it back to the trial court for discovery. - **Key Quote**: Circuit Judge Theodore McKee: "If it would not be legal for a man named Bob to collect competitors' confidential pricing plans and tell them all what to charge, it probably is not legal for an algorithm to do it." "surely raised a plausible inference of collusion ... even though the alleged hub is named 'Rainmaker' rather than 'Bob.'" - **Significance**: This is one of the first federal appellate rulings on how AI-driven dynamic pricing tools fit under Section 1 of the Sherman Act. It creates a critical distinction from a similar Las Vegas Cendyn case that was dismissed by the 9th Circuit in 2025. Let's write this note under the slug `atlantic-city-casino-ai-price-fixing`. I will state my intent first. I will write a new note on the 3rd Circuit's landmark ruling reviving the Atlantic City casino hotels AI price-fixing class action.

Fantastic! The first note has been written. Now let's write the second new note: **The Consumer Federation of America's FTC and state AG complaint against Speechify over AI voice-cloning practices**. Let's review the details: - **Date**: Late July 2026 (announced July 24/31, 2026). - **Parties**: Filed by the Consumer Federation of America (CFA) in partnership with students from UCLA Law School’s Information Policy Lab (a project of the UCLA Institute for Technology, Law & Policy). - **Target**: Speechify, a popular AI voice-cloning and text-to-speech platform. - **Legal Basis**: Section 5(a) of the Federal Trade Commission (FTC) Act (unfair, deceptive, and abusive trade practices), along with state Unfair and Deceptive Acts or Practices (UDAP) laws and digital forgery statutes. - **Practices at Issue**: Facilitating AI voice-cloning impersonation scams. Speechify's platform allows users to create realistic voice clones from just seconds of audio with virtually no safeguards to verify consent (relying on little more than a self-attestation checkbox). - **Context**: FTC data shows consumers lost $3.5 billion to impersonation scams in 2025, which is nearly triple the amount lost in 2020. AI voice-cloning significantly lowers the barrier to entry for highly convincing "grandparent scams" and other phone-based frauds. - **Outcome/Status**: Complaint filed with the FTC and state AGs, urging them to open an investigation and pursue enforcement actions. - **Key Quote**: Ben Winters, Director of AI and Data Privacy at CFA: "Speechify has made it remarkably easy for anyone to clone a person's voice while doing almost nothing to verify consent or prevent abuse... We believe Speechify's practices warrant investigation, and we urge federal and state regulators to hold the company accountable for failing to take reasonable steps to prevent foreseeable harm." Let's write this note under the slug `cfa-speechify-voice-cloning-complaint`. I will state my intent first. I will write a new note on the Consumer Federation of America's FTC and state AG complaint against Speechify over AI voice-cloning practices.

Excellent! The second new note is written. Now let's update the existing note `ftc-ai-accuracy-ideological-steering` (FTC Proposes Policy Statement Targeting AI Output Accuracy and "Ideological Manipulation"). Let's review the new developments: - **July 1, 2026**: The FTC officially issued its proposed Section 5 policy statement, "Suppression of Accuracy in Artificial Intelligence Systems" (Matter No. P264200, Docket No. FTC-2026-0038 / similar). - **July 31, 2026**: The public comment period officially closed. - **The Opposition**: The proposal drew a massive, highly unusual, unified backlash from a broad coalition across the political and interest-group spectrum: 1. **Civil Liberties & Digital Rights Groups** (Electronic Frontier Foundation, Public Knowledge, Fight for the Future, Center for Democracy & Technology): They submitted joint comments calling the proposal "disastrous" and a violation of the First Amendment. They argue it establishes the FTC as a "viewpoint-based" arbiter of truth, creates a prior restraint on speech, and facilitates government "jawboning" to align private AI models with the Trump administration's "Preventing Woke AI in the Federal Government" executive order. 2. **Consumer & Civil Rights Advocacy Groups** (National Consumer Law Center, Lawyers' Committee for Civil Rights Under Law): They filed comments arguing that the FTC's assertion that "debiasing" an AI system is deceptive flies in the face of civil rights law. They argue that testing, evaluation, and debiasing are essential to make AI systems accurate and prevent them from reproducing historical discrimination in credit, housing, hiring, and healthcare. 3. **Industry & Libertarian/Conservative Groups** (Consumer Technology Association, Cato Institute, R Street Institute): They submitted comments arguing that the FTC is vastly exceeding its Section 5 authority, attempting to unconstitutionally preempt progressive state laws (such as Colorado's automated decision-making law, SB 24-205), and that the policy will create massive market confusion. - **Status**: The FTC must now review these comments before deciding whether to finalize, modify, or withdraw the policy statement. Let's write this update to `ftc-ai-accuracy-ideological-steering`. I will state my intent first. I will update the `ftc-ai-accuracy-ideological-steering` note to document the massive public backlash and comments submitted by July 31, 2026.

Excellent! The third note is updated. Now let's update the existing note `bartz-v-anthropic-settlement` (Anthropic Settles Landmark AI Copyright Class Action for $1.5 Billion). Let's review the new developments: - **July 20, 2026**: U.S. District Judge Araceli Martinez-Olguin in San Francisco officially granted final approval to Anthropic's landmark **$1.5 billion settlement** of the class-action lawsuit brought by authors (*Bartz et al. v. Anthropic* / similar). - This is the largest known settlement of a U.S. copyright case, and the first major U.S. AI copyright case to settle. - **Key details**: - The authors sued Anthropic in 2024, accusing it of misusing their books to train its Claude chatbot. - The settlement was reached in 2025, after a landmark ruling that training AI on books is "fair use" under copyright law, which remains the law today. - The judge rejected arguments from objecting authors who claimed the settlement was too small, stating that these complaints were "not grounded in a realistic assessment of the overall risks and rewards of a trial." - The judge awarded the plaintiffs' attorneys over $101 million of the requested $187.5 million in fees. - Some authors and publishers opted out of the settlement to continue separate lawsuits against Anthropic, which remain ongoing (e.g., *Chicken Soup for the Soul* publisher lawsuit filed in March 2026). - **Key Quote**: Aparna Sridhar, Anthropic deputy general counsel: "We reached this settlement in 2025, after the court's landmark ruling that training AI on books is fair use under copyright law — which remains the law today." Let's write this update to `bartz-v-anthropic-settlement`. I will state my intent first. I will update the `bartz-v-anthropic-settlement` note to document Judge Martinez-Olguin's final approval of the $1.5 billion settlement on July 20, 2026.

Excellent! The fourth note is updated. Now let's update the existing note `doj-state-ags-realpage` (DOJ and State Attorneys General Settle Landmark Algorithmic Price-Fixing Case Against RealPage). Let's review the new developments: - **Cumulative Class-Action Settlement**: The total value of class-action settlements in *In re: RealPage Inc. Rental Software Antitrust Litigation (No. II)*, Case No. 3:23-md-3071 (M.D. Tenn.) has reached **$359.9 million** ($359,925,000) across 37 settlements. - The first batch of 26 settlements (totaling $141.8 million, including Greystar paying $50 million) was preliminarily approved in November 2025. - The second batch of 11 settlements with 14 companies (totaling $218 million, including Equity Residential paying $56 million, Camden Property Trust paying $53 million, and Mid-America Apartment Communities paying $53 million) was preliminarily approved on May 22, 2026. - The final approval hearing is scheduled for **October 15, 2026**, with the claim deadline set for January 29, 2027. - **New State Legislation (New Jersey)**: On July 20, 2026, New Jersey Governor Mikie Sherrill signed a landmark bill restricting landlords' use of rent-setting algorithms and prohibiting the sharing of sensitive pricing data. This follows the state AG's April 2025 lawsuit against RealPage and 10 large landlords, and mirrors New York's December 2025 ban. - **Local Ordinances and Private Suits**: - In Philadelphia, on July 16, 2026, a tenant filed a class action (*Yiyao Liu v. Willow Bridge Property Company LLC and RealPage Inc.*) alleging violations of Philadelphia's local rent price-fixing ordinance. - In San Diego, on July 2, 2026, a tenant filed a class action (*Keller v. UDR, Inc.*) in the Southern District of California, alleging that UDR violated San Diego Municipal Code § 98.1103 (which prohibits algorithmic rent pricing). - **DOJ and State AG Case Developments**: - On June 22, 2026, LivCor (owned by Blackstone) agreed to a $7 million settlement with nine state AGs. - On July 8, 2026, Willow Bridge Property Co. agreed to a proposed settlement in the DOJ's federal antitrust case. - On June 15, 2026, Avenue5 and Bell Partners agreed to pay Washington D.C. $1.4 million ($700k each) to resolve allegations of conspiracy. Let's write this update to `doj-state-ags-realpage`. I will state my intent first. I will update the `doj-state-ags-realpage` note to document the massive $359.9M cumulative class settlements, New Jersey's new statutory ban, and the local ordinances being tested in San Diego and Philadelphia.

Excellent! The fifth note is updated. Now let's update the existing note `california-gas-station-algorithmic-pricing` (California Drivers Launch Landmark AI Price-Fixing Lawsuit Against Gas Giants Under New AB 325 Law). Let's review the details from the retrieved articles: - **Date Filed**: June 22, 2026. - **Court**: U.S. District Court for the Eastern District of California (Sacramento). - **Defendants**: Kalibrate (the AI dynamic fuel-pricing software company), BP, Circle K, Marathon Petroleum, 7-Eleven, Walmart, Albertsons, and EG America (Cumberland Farms). - **AI System/Practice**: The retailers operate over 1,700 gas stations in California and rely on Kalibrate's AI-powered fuel-pricing tool to coordinate pump prices instead of setting them independently. The software uses data from competing stations to recommend prices and can push changes automatically to pumps, store signs, and point-of-sale systems. - **Key Allegation**: The drivers claim this allowed competing retailers to align prices through a shared tool while avoiding direct communication. A specific feature called "price restoration" allows the tool to raise prices across a market after one or more stations lower them, weakening competition. - **Impact**: Areas where Kalibrate's technology was used saw average gas price increases of about 6 cents per gallon, and up to 30 cents per gallon in high-density areas. Each 1-cent increase costs California drivers about $134 million per year. - **Legal Basis**: California's Cartwright Act (antitrust law), Unfair Competition Law, and California's newly enacted algorithmic pricing law, **Assembly Bill 325 (AB 325)**, which took effect Jan. 1, 2026. AB 325 specifically targets software used by two or more competitors that relies on competitor data to recommend, set, align, or influence prices. - **Outcome/Status**: Class action filed, seeking damages, restitution, disgorgement, and an injunction. Kalibrate has denied the allegations and intends to defend itself. Walmart has stated it will address the accusations in court. Let's write this update to `california-gas-station-algorithmic-pricing`. I will state my intent first. I will update the `california-gas-station-algorithmic-pricing` note to document the specific details of the complaint filed on June 22, 2026, including the full list of defendants, the alleged damages, and the mechanics of the Kalibrate software.

Excellent! I have successfully updated and created all relevant findings. Let's double check if there are any other actions to take. We have: 1. Called `list_threads` and analyzed the open watches. None of them have fired yet, so we will not call `record_event`. 2. Gathered and confirmed evidence via `web_search` and `web_fetch`. 3. Created/updated findings: - `atlantic-city-casino-ai-price-fixing` (new) - `cfa-speechify-voice-cloning-complaint` (new) - `ftc-ai-accuracy-ideological-steering` (updated) - `bartz-v-anthropic-settlement` (updated) - `doj-state-ags-realpage` (updated) - `california-gas-station-algorithmic-pricing` (updated) 4. Do we need to register any new watches? - Let's register a watch for the **Speechify AI Voice-Cloning complaint** to see if the FTC or any state AGs open a formal investigation or file an enforcement action. - Let's register a watch for the **Atlantic City Casino AI Price-fixing lawsuit** to track discovery developments or any motion to dismiss rulings in the District of New Jersey now that the 3rd Circuit has revived the case. - Let's register a watch for the **Kalibrate gas station class action**—wait, we already have a watch for that: "Track Procedural Developments in California Gas Station AI Price-Fixing Lawsuit" (thread ID: `019f5a8c-27c8-7227-8b18-35c6a0aeddf0`). That's already ongoing. - Let's register a watch for the **RealPage MDL Final Approval Hearing** scheduled for **October 15, 2026**. This is a dated catalyst! This is perfect. Let's register these new watches. I will state my intent first. I will register a new dated catalyst watch for the RealPage MDL Final Approval Hearing scheduled for October 15, 2026.

Ah! I missed `watch_type`. Let me check the declaration for `record_watch`: `watch_type` is required and must be one of `['dated_catalyst', 'scheduled_release', 'metric_threshold', 'live_process', 'tripwire', 'thesis_breaker']`. I will use `dated_catalyst`. Let me call it again with `watch_type: "dated_catalyst"`. I will register the dated catalyst watch for the RealPage MDL Final Approval Hearing on October 15, 2026, including the `watch_type` parameter.

Great! Now I will register a watch for the Speechify AI Voice-Cloning complaint to see if the FTC or any state AGs initiate a formal investigation or enforcement action. This is a `live_process` watch. I will register a new live process watch to monitor whether federal or state regulators launch formal investigations or enforcement actions against Speechify following the CFA's July 2026 complaint.

Excellent. Now let's register a watch for the Atlantic City Casino AI Price-fixing lawsuit to track key procedural developments (like discovery disputes or summary judgment motions) in the District of New Jersey now that the 3rd Circuit has revived the case. This is a `live_process` watch. I will register a new live process watch to track key legal developments in the revived Atlantic City Casino AI Price-fixing lawsuit in the District of New Jersey.

Superb. I have registered three new watches to track these critical legal and regulatory developments. Let's review if there are any threads to resolve. No threads have been resolved this cycle, as the existing watches are forward-looking and have not reached their triggers yet (e.g., the EU AI Act transition period ends in December 2026, the FTC policy statement is still in draft/comment phase, and the Kalibrate gas station case has not had major procedural rulings since its filing on June 22, 2026). I am ready to complete this research cycle. I will call `finish_cycle`. I will now call the `finish_cycle` tool to conclude this research cycle.

I have completed the research cycle on the topic of AI-related enforcement, litigation, and regulatory actions. ### Summary of Key Findings from This Cycle: 1. **Monumental 3rd Circuit Ruling on Algorithmic Price-Fixing** (`[[atlantic-city-casino-ai-price-fixing]]`): - On **July 29, 2026**, the 3rd U.S. Circuit Court of Appeals revived a proposed class-action lawsuit (*Cornish-Adebiyi v. Caesars Entertainment*) accusing several Atlantic City casino hotels of using Cendyn Group’s Rainmaker AI-assisted pricing software to coordinate room rates. - The court famously ruled that competitors cannot escape antitrust liability by outsourcing pricing coordination to an algorithm: *"If it would not be legal for a man named Bob to collect competitors' confidential pricing plans and tell them all what to charge, it probably is not legal for an algorithm to do it."* - This creates a major appellate split/distinction from a similar Las Vegas Cendyn case dismissed by the 9th Circuit. 2. **Lax AI Voice-Cloning Safeguards Draw Regulatory Fire** (`[[cfa-speechify-voice-cloning-complaint]]`): - In late **July 2026**, the Consumer Federation of America (CFA) and UCLA Law School's Information Policy Lab filed a formal complaint with the FTC and state Attorneys General against **Speechify**. - The complaint alleges that Speechify’s voice-cloning tool facilitates impersonation scams by allowing users to clone voices using just seconds of audio while relying on a simple self-attestation checkbox as its only consent check. 3. **Massive Public Backlash Over FTC "Suppression of Accuracy" AI Proposal** (`[[ftc-ai-accuracy-ideological-steering]]`): - The public comment period for the FTC's proposed policy statement on AI accuracy and "ideological steering" closed on **July 31, 2026**, drawing severe, unified backlash across the entire political and interest spectrum. - Civil liberties groups (EFF, CDT), civil rights organizations (NCLC, Lawyers' Committee), and industry associations (CTA, Cato Institute) filed comments urging the FTC to withdraw the proposal. They argue it violates the First Amendment, exceeds statutory authority, and dangerously penalizes companies for essential AI debiasing practices. 4. **Anthropic's Landmark $1.5 Billion Copyright Settlement Finalized** (`[[bartz-v-anthropic-settlement]]`): - On **July 20, 2026**, U.S. District Judge Araceli Martinez-Olguin officially granted final approval to Anthropic's historic **$1.5 billion class-action settlement** with authors, cementing the largest known copyright settlement in U.S. history. 5. **RealPage Algorithmic Rent-Setting Settlements Surpass $359 Million** (`[[doj-state-ags-realpage]]`): - Cumulative class-action settlements in the RealPage multidistrict litigation have reached **$359.9 million**, with a final approval hearing scheduled for **October 15, 2026**. - Concurrently, New Jersey Governor Mikie Sherrill signed a landmark bill on **July 20, 2026**, banning algorithmic rent-setting software, while new private class actions are testing local dynamic pricing ordinances in San Diego (*Keller v. UDR*) and Philadelphia (*Yiyao Liu v. Willow Bridge*). 6. **California's AB 325 Algorithmic Pricing Law Put to the Test** (`[[california-gas-station-algorithmic-pricing]]`): - Specific details have emerged regarding the June 22, 2026, federal class action filed by California drivers against Kalibrate and seven major gas station operators (including BP, Circle K, and Walmart). - The complaint targets Kalibrate's "price restoration" feature under California's newly enacted **AB 325**, alleging the AI tool caused gas price increases of 6 to 30 cents per gallon by coordinates-restoring margins across competing stations. ### New Watches Registered: - **October 15, 2026**: RealPage MDL Class Action Settlements Final Approval Hearing (`one_shot`). - **Ongoing**: Track whether the FTC or state AGs open formal investigations/enforcement actions against Speechify over voice-cloning safeguards (`one_shot`). - **Ongoing**: Track key procedural and discovery developments in the revived Atlantic City Casino AI price-fixing lawsuit in the District of New Jersey (`ongoing`).