Anthropic

Independent Sources

Shown for context — not merged with USAComplaints' own numbers.

Contacts & Login

• Official website: https://www.anthropic.com/
• Claude product site: https://claude.com/
• Help Center (billing, refunds, appeals): https://support.claude.com/en/
• Claude account login (Settings > Billing to cancel; Get help to reach support): https://claude.ai/
• Account and cancellation email named in the Consumer Terms: [email protected]
• Privacy requests: [email protected]
• Data Protection Officer: [email protected]
• Safeguards warnings (email named on the Help Center page): [email protected]
• Service status page: https://status.claude.com/
• Privacy Center: https://privacy.claude.com/en/
• BBB-listed number (not published by Anthropic as a support line): (415) 571-3346

Locations

Using Claude or the Anthropic API? Billing, cancellation, refunds, limits and appeals

Anthropic PBC sells Claude to individuals under its Consumer Terms of Service (the U.S. and rest-of-world version of the page shows "Effective October 8, 2025") and sells the Claude API and Console to developers and businesses under separate Commercial Terms (effective June 17, 2025). What follows is what Anthropic's own pages say on the questions readers most often have, as they read on September 30, 2026; the terms can change, so check the live pages before relying on them.

• Cancelling Pro or Max. Anthropic's Help Center says: "You may cancel your Pro or Max plan subscription at any time ... Your cancellation will take effect at the end of your current billing period and you can continue using your paid plan until then." On claude.ai or Claude Desktop the path is Settings > Billing > Cancel. Subscriptions bought in the iOS or Android app are cancelled in the Apple or Google account instead. The Help Center adds: "To avoid being charged for your next billing period, cancel your subscription at least 24 hours before your next billing date."
• Refunds. The Consumer Terms say: "Except as expressly provided in these Terms or where required by law, all payments are non-refundable." The Help Center's refund page repeats this and says: "To request a refund for a payment defined as refundable in our Terms of Service, follow these instructions": log in, click your initials or name, choose "Get help", click "Send us a message", then "Accept" and "Claude Refund Request", select a refund reason and "Follow the prompts to check your refund eligibility"; the last step reads "If eligible, continue in the support messenger to cancel and refund your plan." The same page describes other cases: when a Pro or Max subscriber upgrades to a Team plan, "your individual subscription is cancelled automatically and you receive a prorated refund for the unused portion of your billing period" and "Any prepaid usage credits on your individual account are refunded"; for plans bought in Claude for iOS, "only Apple can review and issue a refund" (Apple's route is "I'd like to" then "Request a refund"), and the page adds: "Requesting a refund doesn't cancel your subscription. To avoid future charges, you'll also need to cancel."; for Claude for Android, Anthropic's support team "will need to check your eligibility". It also says support "cannot process refunds during pending disputes" with a bank. In the European Economic Area and United Kingdom the page describes a refund within 14 days of purchase, prorated to use; the U.S. Consumer Terms text I read contains a 7-day cancellation right only for residents of Brazil, Mexico, South Korea and Taiwan.
• Auto-renewal, price changes and termination. The Terms say paid subscriptions renew automatically "until you cancel"; fee increases require notice "at least 30 days in advance"; if Anthropic terminates a subscriber for a Terms violation "you will not be entitled to any refund", but if it ends a subscription for any other reason and it was bought on its website it will "refund you, on a pro rata basis".
• Usage limits. Anthropic's Max plan page says "Max 5x includes five times the Pro plan's per-session usage allowance" and "Max 20x includes 20 times", that "Your session-based usage limit will reset every five hours", and that Max plans also have a weekly usage limit "that applies across all models", which "resets at a fixed time each week that is assigned to your account"; "You can see your next reset time in Settings > Usage." It adds: "to manage capacity and ensure fair access to all users, we may limit your usage in other ways, such as weekly and monthly caps or model and feature usage, at our discretion." A separate Help Center article says limit resets "are given occasionally to eligible plans". The Help Center pages I read on 2026-09-30 do not describe a process for support to restore usage manually. Readers preparing a usage or billing question for support may find it useful to keep screenshots of Settings > Usage showing the reset times, the model and app used, timestamps, receipts, and the status-page entry (status.claude.com) for the same period. Two pending class actions (see the litigation section: Kahn and Smith; In re Claude Service Degradation Litigation) challenge how usage was delivered; Anthropic's position is in its court filings, which I did not read in full. Optional "usage credits" are billed separately at API rates, are prepaid, can be capped monthly, and can be switched off in Settings > Usage.
• Reaching support. The Help Center states: "We don't offer phone support." and that support is asynchronous. Its "How to get support" page says Pro and Max plan users, Team and Enterprise plan owners and Console Admins have access to Fin (the AI support bot) and "Further assistance from our Product Support team": log in, choose "Get help", "Send us a message", and if the issue needs investigation "Fin will pass your inquiry along"; replies arrive in the same conversation and by email. For Team plan and Console non-owners, the page says "Human specialist support is not directly available for your account seat type" and the Primary Owner, Owner or Console Admin can reach out on their behalf. The page also says: "Keep one conversation per issue. Opening a second request for the same problem creates a duplicate that has to be re-triaged, which usually slows the answer down. If you haven't heard back, reply in your existing conversation instead." If you cannot log in, it lists account deletion, data exports and subscription support (help locating login information, cancellations, refund-eligibility checks) and says to start a conversation from the Help Center and select "I can't login."
• Account suspended or banned. The Help Center page "Safeguards warnings and appeals" says: "If you feel your account has been wrongly suspended or terminated, please go to claude.ai, log in with your banned account, and fill out the appeal form" and "You must be logged in to access the appeal form." For an account banned for violating the Usage Policy, the page says the holder can still log in at claude.ai to export data or delete the account without contacting Anthropic (the options appear on the same screen as the appeal form), that "All banned users can delete their own account", and that self-serve exports are available to all banned users but "the data you’re able to access and export may be restricted depending upon the Usage Policy violation". If an organization you belong to is on hold, the page says to click "Request a review" on the affected organization. For warnings, the same page (checked on 2026-09-30) says: "If you believe we've made a mistake in issuing you a warning, please email [email protected] with the details of your situation and your account information." It also carries a note that response times are "currently longer than normal".
• Deleting your data or account. The Help Center's "Delete your Claude account" article (dated August 19, 2026) says: "Deleting your account is permanent" and advises exporting data first. For a paid Pro or Max plan it lists three steps: cancel from Billing settings, "Wait until the end of your current subscription period", and delete the account once the subscription lapses (Settings > Account > Delete account on the web). It notes that in some scenarios you "need to contact our team to delete your account". The Privacy Policy says individual deleted conversations "will be removed immediately from your conversation history and automatically deleted from our back-end within 30 days"; privacy rights requests go to [email protected].
• Privacy choices. Consumer accounts have a model-improvement setting; see the background and FAQ sections for the exact wording and retention periods. Privacy requests go to [email protected] (data protection officer: [email protected]).

Other places to raise a problem, none of them run by Anthropic: your card issuer or app store for a billing dispute, the Better Business Bureau (which lists a profile for Anthropic), the Federal Trade Commission at ReportFraud.ftc.gov, and your state attorney general's consumer-protection office.

Regulator and federal-government matters

I did not find any Federal Trade Commission, state attorney general or state privacy-regulator enforcement action against Anthropic, or a consent order, in the sources I checked on 2026-09-30. That is a scoped result: the FTC's and California Attorney General's site-search pages did not return filtered results in my text tools, so this is not a complete regulator search.

The government matter that has produced court decisions is a federal procurement dispute with the U.S. Department of War (the name used in the court captions). Two courts have issued decisions in separate proceedings, described below with the court, statute, relief and posture each states:

• U.S. District Court, N.D. California (Judge Rita F. Lin), No. 3:26-cv-01996-RFL, August 27, 2026. The opinion states that "On February 27 and March 3, 2026, President Trump and Secretary of War Hegseth designated Anthropic a supply chain risk to national security, ordered all federal agencies to permanently stop using Anthropic's products, and banned all defense contractors from doing any business with Anthropic." The court granted Anthropic's summary-judgment motion in part and the government's in part. Its Order of Final Relief declares that the challenged actions violate the First Amendment ("unlawful retaliation") and the Fifth Amendment's Due Process Clause, permanently enjoins the defendants (other than certain "Non-Participating Defendants") from implementing them, and directs them to rescind related guidance; the court ruled for the defendants on an ultra vires (separation of powers) claim and on claims against some agencies. A judgment was entered the same day.
• U.S. Court of Appeals, D.C. Circuit, No. 26-1049 (consolidated with 26-1162), September 25, 2026. A three-judge panel (Henderson, Katsas, Rao) entered a per curiam judgment that Anthropic's "petitions for review be denied", by a 2-1 vote: the opinion for the court is by Judge Katsas and Judge Henderson dissented. The opinion describes the case as arising from the Department's decision to exclude Claude from its supply chain under the Federal Acquisition Supply Chain Security Act of 2018 after Anthropic "refused to relax contractual prohibitions on the use of Claude for lethal autonomous warfare or domestic surveillance", and says: "We reject these challenges." The majority reads the definition of "supply chain risk" in section 4713 as much broader than the one in section 3252 and as not requiring bad motive; it says it has "no quarrel" with the Northern District's conclusions that section 3252 requires bad motive and that Anthropic acted with none. The dissent would read the section 4713 definition more narrowly. By a separate order the same day the court directed the clerk to withhold the mandate until seven days after disposition of any timely petition for rehearing; the RECAP docket entries I read on September 30, 2026 show no later entry.

The two decisions concern different statutory designations as the courts describe them (10 U.S.C. § 3252 in the district court's opinion; 41 U.S.C. § 4713 in the D.C. Circuit's), and the D.C. Circuit's opinion says of the district court's decision: "the issues flagged by Anthropic are not the same in both cases." A separate Ninth Circuit appeal by the government defendants (No. 26-2011, opened April 2, 2026) was stayed by an order of April 27, 2026; I did not find a later entry in that docket, and I did not review any filings made after the two decisions above. Neither decision concerns the terms of consumer or API subscriptions.

Company background, prices and data terms

Who operates Claude. Anthropic PBC's Privacy Policy (effective September 10, 2026) gives its registered address as 548 Market St, PMB 90375, San Francisco, CA 94104. In a complaint Anthropic itself filed on July 1, 2026 (Anthropic, PBC v. Abnormal AI, Inc., N.D. Cal.), it describes itself as "a public benefit corporation organized under the laws of the State of Delaware and headquartered in San Francisco, California" and says it "was founded in 2021". Its company page says: "Anthropic is an AI safety and research company." For users in the European Economic Area, the UK and Switzerland the Privacy Policy names Anthropic Ireland, Limited (Dublin) as data controller; the claude.com footer reads "Services in the EU are provided by Anthropic Ireland Limited."

Ownership statement. In a corporate-disclosure filing on September 28, 2026 in Sony Music Publishing (US) LLC v. Anthropic PBC (N.D. Cal. 5:26-cv-09217), Anthropic identified "No Corporate Parent" and listed as "Other Affiliate" Alphabet, Inc., Google LLC, Amazon.com, Inc. and Amazon Web Services, Inc. The filing, as shown on the docket, does not describe the nature of those affiliations.

Products and prices (claude.com/pricing, September 30, 2026). Free ($0); Pro ($17 per month with an annual subscription, $200 billed up front, or $20 if billed monthly); Max, "From $100" per month, with 5x or 20x more usage than Pro (the Help Center lists Max 5x at $100 and Max 20x at $200 per month for web subscriptions); Team and Enterprise plans; and the Claude API, priced per token. The page notes "Prices shown don't include applicable tax. Price and plans are subject to change at Anthropic's discretion." Pro and Max include Claude Code. The Help Center explains that a Claude subscription does not include API or Console access, which is billed separately.

API and Console billing. The Commercial Terms say Anthropic "may require prepayment for the Services in the form of credits", subject to its Supplemental Credit Terms (effective March 4, 2024). Those terms state that credits "are non-refundable", that usage credits "expire one calendar year from the date Anthropic sends the Confirmation Notice", that they expire if the account is closed, and that if a customer turns on auto-reload it "authorizes Anthropic to charge the payment method ... on a recurring basis ... until Customer opts out". Under the Commercial Terms disputes go first to good-faith negotiation and then to binding arbitration (JAMS in San Francisco for customers outside the EEA, Switzerland and UK), with a class-action waiver. The consumer terms I read for Claude.ai are different: they choose California law and "state or federal courts located in San Francisco" and contain no arbitration clause in the text I read.

Privacy and data terms (Anthropic's wording). The Consumer Terms state: "We may use Materials to provide, maintain, and improve the Services ... including training our models, unless you opt out of training through your account settings. Even if you opt out, we will use Materials for model training when: (1) you provide Feedback ... or (2) your Materials are flagged for safety review." Anthropic's August 28, 2025 announcement said existing consumer users had until October 8, 2025 to accept the updated terms and choose the setting and that data may be kept for five years if training is allowed. Its Privacy Center says chats used for model improvement may be retained in de-identified form for up to five years; feedback data is kept five years; inputs and outputs flagged as Usage Policy violations are kept up to two years, with classification scores up to seven years; and Incognito chats are not used to improve Claude. The Privacy Policy says deleted conversations are removed from the back end "within 30 days", that Anthropic does not "sell" personal data as defined by law, and that it may share data with authorities where it has "a good-faith belief that disclosure is reasonably necessary" for reasons the policy lists (legal process, preventing serious harm, fraud, or enforcing its terms). Settings > Privacy on claude.ai offers a data export; the export link expires after 24 hours.

Name confusion. A search of the SEC's EDGAR company list for "anthropic" on September 30, 2026 returned 35 registrants whose names contain "Anthropic"; the names shown (for example "... a Series of CGF2021 LLC") are those of investment vehicles, and no registrant named Anthropic PBC itself appeared on the list. I did not open those filings.

Complaint and review records, reported separately

Sources that record complaints or reviews about Anthropic are reported separately here and are not combined or averaged:

• USAComplaints: 0 published complaints about Anthropic or Claude were found in this site's database as of September 30, 2026.
• BBB profile listed as "Anthropic" (Marketing Software), 548 Market St, PMB 90375, San Francisco, website www.anthropic.com (BBB profile opened 2026-09-30): BBB says "This business is NOT BBB Accredited." Rating: D-; the reason BBB shows is "Failure to respond to 263 complaint(s) filed against business". Complaints tab: "270 total complaints in the last 3 years" and "248 complaints closed in the last 12 months" (BBB states that the profile "includes complaints for Anthropic's headquarters and its corporate-owned locations"). Reviews tab: 1.15 out of 5, "Average of 27 Customer Reviews". BBB also lists "BBB File Opened" August 26, 2023 and "Business Started" October 4, 1998 (raw profile data) with "Years in Business: 28"; these BBB-listed dates do not match the 2021 founding year Anthropic PBC states in its own court filing (see company background), and BBB does not say which entity its business-start date refers to. The address and website match those Anthropic PBC publishes (Privacy Policy address; anthropic.com), but BBB does not say the profile is the legal entity's file. BBB says it does not verify the accuracy of information provided by third parties. These BBB figures are BBB's own and are not combined with Trustpilot or any other source.
• Trustpilot, two profiles, as displayed on 2026-09-30 (counts change): claude.ai, 1.5 from 2,011 reviews, and anthropic.com, 1.3 from 491 reviews. Both are labelled "Unclaimed profile" and carry the notice "No history of asking for reviews. This company hasn't invited their customers, so reviews may not be representative." Trustpilot's own AI-generated summaries on those pages describe reviewers mentioning billing, refunds, support and usage limits; I have not counted or verified individual reviews.
• Regulator complaint databases: the CFPB database covers financial products and was not used; no other regulator complaint count was located.

Service issues Anthropic has itself published. On April 23, 2026 Anthropic published an engineering post, "An update on recent Claude Code quality reports", saying reports that responses had worsened were traced to three changes affecting Claude Code, the Claude Agent SDK and Claude Cowork (a default reasoning-effort change on March 4, reverted April 7; a caching change on March 26 with a bug fixed April 10; and a system-prompt change on April 16, reverted April 20). It states "We never intentionally degrade our models" and "As of April 23, we're resetting usage limits for all subscribers." The post says the API was not affected. Anthropic's status page (status.claude.com) showed "All Systems Operational" on September 30, 2026 and listed an incident on September 29, 2026 with impact from 07:00 to 07:59 PT during which "Some messages sent between 14:00 and 14:59 UTC may not have been saved" and sign-in, new chats, purchases and file uploads were affected, plus incidents on September 22 (elevated errors on several models) and September 16 (Google Play subscriptions). A text render of the status page showed 90-day uptime of 99.42 percent for claude.ai and 99.52 percent for the API (Anthropic's own figures).

Lawsuits naming Anthropic, with procedural status

Court filings are allegations unless a court has ruled; the notes below give each matter's procedural status from the dockets and orders I read on September 30, 2026 (CourtListener/RECAP and the court documents). CourtListener/RECAP mirrors PACER and may not show every filing or the text of every document; where I write that I found no ruling, it means none appeared in the entries and documents I reviewed through the date stated, not that PACER itself was searched.

Authors' copyright class action: Bartz v. Anthropic PBC (N.D. Cal. No. 3:24-cv-05417-AMO). The court-authorized settlement notice says the suit alleged that Anthropic infringed copyrights by downloading datasets containing copyrighted books (from Library Genesis and Pirate Library Mirror), and that "Anthropic denies all the allegations and denies that it did anything wrong" and argues its use was fair use. It also says that in June 2025 the district court ruled that using books to train AI models was fair use but left for trial whether downloading and using works from those two sources was. The settlement is $1.5 billion, non-reversionary. On July 20, 2026 Judge Araceli Martinez-Olguin granted final approval, approved the plan of allocation, overruled all objections, awarded class counsel $101,561,111 in fees (with 10 percent held back until a post-distribution accounting), $2,635,197.46 in expenses and $15,000 service awards to each of three class representatives, dismissed the action with prejudice and entered final judgment while reserving jurisdiction over implementation. The order records 350 valid opt-outs covering 1,802 works and that 440,490 of the 482,460 works on the Works List had been claimed as of April 16, 2026. Class counsel's status report of September 2, 2026 says two notices of appeal were filed by law firms excluded from the fee award (Edelson PC and Oppenheim + Zebrak, LLP; and Cowan DeBaets Abrahams & Sheppard LLP; Ninth Circuit Nos. 26-5391 and 26-5455), that both relate only to the fee award so the parties agree the settlement's Effective Date has passed, and that an objector's filing styled as a request to appeal or intervene was not a notice of appeal. The report says distributions should start by early November and no later than November 15, 2026, and that after deductions each work is allotted approximately $2,203.56 of the funds then in the settlement fund (449,731 claimed works; about $991 million available for this distribution), subject to adjustments for further interest or accepted late claims; the figure does not include the final $450 million installment (plus interest), which is due the earlier of September 25, 2027 or 30 days after a qualifying Anthropic funding event. For comparison, the settlement website's FAQ (read on 2026-09-30) describes the settlement as providing "approximately $3,000 per work" plus interest, less the court-approved costs and fees; the FAQ text I read does not restate the $2,203.56 figure. On payment timing, the settlement website's homepage text, which the court approved by order of September 15, 2026 (Dkt. 691-1 and 692), says the parties agree the Effective Date has passed "assuming that no late-filed appeal is permitted to go forward" and: "We currently expect to send the first round of payments on or before November 15, 2026 for works (a) on which claimants do not disagree on the appropriate allocation (or who quickly reach agreement to resolve any disagreement) and (b) for which claimants have provided all the information required to process their payments." For other works, it says it "may take some time to resolve disagreements and collect additional information before payments can be issued" and that payments otherwise due are held in reserve in escrow. A further Notice (Other) was filed on September 29, 2026 (Dkt. 693), which I did not review. The settlement website says the claim, opt-out and objection deadlines (March 30 and February 9, 2026) have passed. For claimants with a timely claim, it directs them to an Allocation Portal (using the Claim Number and PIN in their notice) and says co-claimants who disagree on allocation should submit information "within 60 days of receiving access to the consolidated claim information", a period that "does not create a deadline by which all disagreements between co-claimants must be resolved". Administrator contact: 877-206-2314 or [email protected]. Who is in the class: the website says "You may be a Class Member if you are the legal or beneficial copyright owner of a book included in the Anthropic Copyright Settlement", and its FAQ says the only way to determine this is the Works List and "If you are the legal or beneficial owner of a work that is in the Works List, then you are a Class Member." The eligibility test as the website states it is ownership of a listed work; it does not mention use of Claude.

Music publishers. (1) Concord Music Group, Inc. et al. v. Anthropic PBC, N.D. Cal. No. 5:24-cv-03811-EKL (the docket shows an earlier Middle District of Tennessee case with the same caption closed June 24, 2024): the publishers describe it as a suit over Anthropic's use of 499 of their musical compositions "as the input and output of certain of Anthropic's Claude AI models". As of September 22, 2026 the docket shows summary-judgment and expert-exclusion motions with reply briefs filed that day; a July 15, 2026 hearing was affected by an April 24, 2026 order continuing case deadlines about 90 days, and I found no ruling on those motions. (2) Concord Music Group v. Anthropic PBC (N.D. Cal. No. 5:26-cv-00880-EKL, filed January 28, 2026): a second suit alleging that Anthropic torrented pirated books containing 714 of the publishers' compositions; the amended complaint filed May 21, 2026 also names Dario Amodei and Benjamin Mann. Anthropic and Mr. Amodei have partial motions to dismiss that were fully briefed on September 14, 2026; on September 21, 2026 the court granted stipulations to coordinate it with the BMG case. (3) BMG Rights Management (US) LLC v. Anthropic PBC (No. 5:26-cv-02334-EKL, filed March 17, 2026) alleges copying of BMG lyrics for training and in Claude outputs. (4) Round Hill Music LP et al. v. Anthropic PBC (No. 5:26-cv-08505, filed August 17, 2026) and (5) Sony Music Publishing (US) LLC, Warner Chappell Music, Inc. et al. v. Anthropic PBC, Dario Amodei and Benjamin Mann (No. 5:26-cv-09217, filed August 28, 2026) were newly filed and, in the entries I read, had only appearance and disclosure filings.

Reddit, Inc. v. Anthropic PBC. Reddit sued in San Francisco Superior Court (No. CGC-25-625892) alleging that Anthropic used unauthorized automated access to scrape Reddit content to train Claude, pleading breach of contract, unjust enrichment, trespass to chattels, tortious interference and unfair competition. Anthropic removed the case to federal court on July 3, 2025 (N.D. Cal. No. 3:25-cv-05643); on March 30, 2026 the court granted Reddit's motion to remand to state court. I did not review the state-court docket.

Class actions by Claude subscribers. (1) Kahn and Smith v. Anthropic PBC (consolidated), N.D. Cal. No. 3:26-cv-05763-AGT (Magistrate Judge Alex G. Tse): Kahn was filed June 14, 2026 and Smith (No. 26-cv-07789) on July 27, 2026; the court found them related on August 7, 2026, and after a joint stipulation the cases were consolidated by order of September 4, 2026. The Consolidated Complaint was filed September 8, 2026. The original Kahn complaint alleged that Max 20x delivers "just six to eight times the usage of Pro" and Max 5x "just three-and-a-half times", and that a claimed "50% savings" is false, for a proposed class of U.S. persons who bought Max 5x or 20x on Claude.com or the desktop app from April 9, 2025. Anthropic's motion to dismiss the Consolidated Complaint (filed September 22, 2026; hearing set for November 6, 2026) describes it as asserting six California-law claims (Consumers Legal Remedies Act, False Advertising Law, Unfair Competition Law, negligent misrepresentation, breach of contract and breach of the implied covenant), quotes it as asserting that Anthropic promised "5 times more usage . . . [or] 20 times more usage than Pro on a monthly (or at least weekly) basis", and argues that "Anthropic never promised a fixed total amount of usage or capacity" and that its statements refer to "per session" usage compared with Pro. These are the parties' positions, not findings; no ruling has been made. (2) In re Claude Service Degradation Litigation, N.D. Cal. No. 3:26-cv-07699-RFL (Judge Rita F. Lin): Pascual (filed July 24, 2026) and Patel/Morgan (filed July 28, 2026) were consolidated on August 26, 2026 and a Consolidated Class Action Complaint was filed September 14, 2026; a September 11, 2026 order gives Anthropic thirty days after that filing (that is, until October 14, 2026) to answer, move or otherwise respond, and an initial case management conference is set for November 4, 2026. The Patel complaint alleges that a March 2026 peak-hour change and March-April 2026 Claude Code defects made paid usage limits deplete faster while prices stayed the same, that no automatic pro rata credit or refund was provided, and pleads unfair-competition, false-advertising, Consumers Legal Remedies Act, contract, implied-covenant and unjust-enrichment claims for a class period beginning March 4, 2026; Anthropic's April 23, 2026 post (above) acknowledges the Claude Code changes and reset usage limits. I did not read the Consolidated Complaint of September 14 itself.

Antitrust class action covering four companies: Buist et al. v. Anthropic, PBC, OpenAI OpCo, LLC, SpaceXAI LLC and Google LLC, N.D. Cal. No. 3:26-cv-10693 (filed September 18, 2026; Magistrate Judge Nathanael M. Cousins). This is a separate multi-defendant matter, not a Claude usage-limit or billing suit: a putative nationwide class of U.S. purchasers of paid individual subscriptions to ChatGPT, Claude, Grok or Gemini alleges an agreement among the four companies to slow the pace of model improvement in violation of Section 1 of the Sherman Act, relying on public statements the complaint attributes to company executives on and after September 12, 2026. The docket shows Anthropic was served on September 24, 2026 with an answer date of October 15, 2026 and an initial case management conference set for December 23, 2026; no response, class certification or ruling appears in the entries I read, and I did not independently confirm the statements the complaint quotes. These are unproven allegations.

Other copyright suits by authors and publishers (separate from the Bartz class): Carreyrou et al. v. Anthropic PBC et al. (N.D. Cal. No. 5:25-cv-10897, filed December 22, 2025; the original complaint also named Google, OpenAI, Meta, xAI and Perplexity) - Anthropic filed a consolidated motion to dismiss on September 24, 2026 with a hearing set for December 17, 2026; and pending N.D. Cal. cases Kwon (5:26-cv-04649), Cruz (3:26-cv-04482), Shakespeare (5:26-cv-05931), Cognella, Inc. (4:26-cv-04056) and Chicken Soup for the Soul, LLC (5:26-cv-04218), plus Gilbert v. Anthropic PBC (D. Mass. 3:26-cv-30081, with an Anthropic counterclaim at issue). I read only the introductions and docket status for these.

Other matters where Anthropic is a party include Anthropic's own trademark suit Anthropic, PBC v. Abnormal AI, Inc. (N.D. Cal. 3:26-cv-06754, filed July 1, 2026); a patent suit, University of Tennessee Research Foundation v. Anthropic PBC (D. Del. 1:26-cv-00887, filed July 20, 2026); and several self-represented or individual suits with docket-only information available to me, for example Grossinger (E.D. Pa. 2:26-cv-04705, filed July 3, 2026, listed on the docket as an Americans with Disabilities Act case; dismissed with prejudice on August 7, 2026 in a memorandum and order granting leave to proceed in forma pauperis; a notice of appeal was filed September 9 and docketed in the Third Circuit as No. 26-3299 on September 15, 2026), Haugland (D. Haw. 1:26-cv-00382 and D. Del. 1:26-cv-00997), Gribble (D. Haw. 1:26-cv-00426), Jiang (W.D. Pa. 3:26-cv-01839), Brunell (D. Vt. 2:26-cv-00280) and Williams (S.D.N.Y. 1:26-cv-01566, multiple defendants). I did not review the allegations in these.

Is Anthropic Legitimate?

Anthropic PBC is identified in its own filings as a Delaware public benefit corporation headquartered in San Francisco, founded in 2021, with a registered address at 548 Market St, PMB 90375, San Francisco, CA 94104 per its Privacy Policy. It publishes Consumer Terms, Commercial Terms, a Usage Policy and a Privacy Policy, runs a public status page and Help Center, and states it does not offer phone support. Anthropic told a federal court on September 28, 2026 that it has no corporate parent and lists Alphabet, Google and Amazon entities as affiliates.

Independent records are mixed and are reported without a score of this site's own: BBB shows a not-accredited profile listed as "Anthropic" with a D- rating (reason shown: "Failure to respond to 263 complaint(s) filed against business"), a complaints tab reading "270 total complaints in the last 3 years", and a 1.15 out of 5 review average from 27 reviews, and Trustpilot's two unclaimed profiles show 1.5 (2,011 reviews) and 1.3 (491 reviews), both with a notice that reviews may not be representative. Court records show an authors' class settlement of $1.5 billion with final approval and judgment entered, several pending copyright suits, two pending class actions filed in 2026 by Claude subscribers about plan marketing and usage limits (allegations unproven; Anthropic has moved to dismiss the consolidated Kahn and Smith complaint), a separate pending antitrust class action against four AI companies including Anthropic, and two federal-court decisions in the Department of War dispute, made under different statutes. This page assigns Anthropic no rating and no label; the sources above are provided so readers can weigh them.

Court & Public Records

Company Relationships

• Affiliated with: Alphabet, Inc.
• Affiliated with: Google LLC
• Affiliated with: Amazon.com, Inc.
• Affiliated with: Amazon Web Services, Inc.
• Affiliated with: Anthropic Ireland, Limited

Frequently Asked Questions