We’re currently waiting for Judge Araceli Martinez-Olguin to issue final approval of the historic settlement in the Bartz v, Anthropic copyright infringement lawsuit. (If you need a refresher, my past blog posts include a general overview of the settlement and an April update.)

One of the interesting outcomes–so far–has been that while participation by class members has been extraordinarily robust (91.3%, with claims filed for 440,490 of the 482,460 works on the elgible works list), the number of class members who opted out has been quite small–just 350, according to a motion for final approval filed by plaintiffs in March (the opt-out deadline was February 9).

Opting out preserved those individuals’ right to sue Anthropic on their own. And indeed, in recent months, several lawsuits have been filed, representing approximately a third of the writers who opted out. Each suit is interesting in its own way.

Carreyrou v. Anthropic

If you’ve been following the progress of the settlement, you may remember a group called ClaimsHero. A law firm whose leadership team members do not appear to have law degrees, ClaimsHero got in trouble last year with the court over its attempts to recruit Anthropic class members for representation in alternative copyright lawsuits, using the lure of a higher payout (since the settlement will pay around $3,000 but the statutory maximum for copyright infringement is $150,000) without adequately disclosing to potential clients that signing with ClaimsHero meant authorizing ClaimsHero to opt them out of the settlement. (I wrote about this at the time.)

ClaimsHero was forced to change its recruitment and advertising tactics. However, it continued to seek clients, and in December 2025, its efforts bore fruit: six authors, including author and journalist John Carreyrou, filed suits against Anthropic and several other AI companies, based on allegations similar to those that led to the settlement they’d opted out of: that the AI companies had infringed their copyrights by copying and storing pirated books in order to use them for AI training.

The suits, which seek a judgment of willful infringement, statutory damages of up to $150,000 per infringed work, and attorney fees and expenses, are being litigated not by ClaimsHero (one of the things that came out when ClaimsHero was hauled into court was that it had no litigation experience) but by a partner law firm. You’ll also notice that I said suits, plural. The Carreyrou filing isn’t a class action, but six individual lawsuits related by a common cause of action and filed together.

Class action settlements, the plaintiffs argue, serve defendants rather than creators, sidestepping creators’ right to have statutory damages individually determined and allowing companies like Anthropic to “easily extinguish thousands upon thousands of high-value claims at bargain-basement rates, eliding what should be the true cost of their massive willful infringement.”

This unusual approach, with individual cases joined together into a single action, is used by third-party claims filers, companies that specialize in recruiting claimants in class action settlements and batch-filing individual claims on their behalf. The tactic doesn’t always pan out, and whether it will succeed here remains to be seen. The plaintiffs have already suffered a setback, with most of the AI companies named in the original complaint severed from the case. The case has also been consolidated with several other infringement cases against Anthropic.

Back to ClaimsHero for a minute…ClaimsHero’s recruitment efforts have been aggressive. Between February and April, I got at least seven of these emails.

“Potentially” in that first bullet point is doing a lot of work; as far as I know, the Carreyrou action is the only one ClaimsHero has facilitated to date. And although I haven’t gotten a solicitation in a while, ClaimsHero is still recruiting–and there’s still some, well, lack of transparency. Here’s its current website pitch:

But wait, you’re thinking, where is that $750k per work figure coming from? Isn’t the $150k the maximum for statutory damages in a copyright suit? Well, yes, and ClaimsHero does clarify that, but only if you bother to scan down the page to find the asterisk:

This is the same kind of bait-and-switch advertising as calorie claims on food packaging that don’t mention that the calorie count is per serving.

Daniel Benjamin Gilbert v. Anthropic

This second infringement case filed by an author who opted out of the Anthropic settlement touches on a sore issue for writers who did not opt out: they must split their payouts with their publishers if their contracts are still in force and the publisher files a claim.

Originally filed in May and amended on July 7, Gilbert’s complaint makes familiar allegations (infringement based on Anthropic’s copying and storage of pirated books) and seeks familiar relief (including full statutory damages, attorney fees and costs, and destruction of the infringed material). However, Gilbert also wants to make sure his publisher, Wiley, can’t share of any of the proceeds.

Gilbert says that he notified Wiley in writing of Anthropic’s infringement, and “asked Wiley to advise within thirty days whether it intended to pursue the claim, and stated that if Wiley did not, he would proceed as provided in Paragraph 16.” Wiley did not respond within that time. Later, however, it notified Gilbert that it had submitted a claim under the settlement–but that the claim was moot thanks to Gilbert’s opt-out (under the terms of the settlement, an opt-out by one rightsholder opts out all other rightsholders as well).

The question here is whether, should Gilbert win his case, Wiley could claim a share of the recovery by arguing that making a claim under the settlement counted under Paragraph 16, despite the opt-out issue and even though it took more than 30 days to respond to Gilbert’s notification. Gilbert argues that it can’t: “Because Wiley declined to pursue the claim after Plaintiff’s request, as between Plaintiff and Wiley any recovery in this action, which Plaintiff prosecutes at his own expense, belongs solely to the Author, and Wiley has no contractual entitlement to share in it.”

Whether a court will agree remains to be seen. It’ll be interesting to watch how this particular case unfolds.

Shakespeare v. Anthropic

Filed June 17, this is the most recent infringement suit by Anthropic settlement opt-outs. Defendants are not just Anthropic, but Dario Amodei, Anthropic’s co-founder and CEO, and Benjamin Mann, who is identified as “a founder and member of the technical staff of Anthropic” and who the suit alleges personally handled the downloading of the pirate datasets that included the plaintffs’ books.

Plaintiffs bring this case to address Defendants’ unlawful downloading and exploitation of their works from library websites known to consist of pirated works by using BitTorrent, a file-sharing technology widely used for mass copyright infringement. Defendants downloaded works by torrenting an enormous number of unauthorized copies of Plaintiffs’ works from illegal shadow libraries to avoid paying for those works, while at the same time uploading via torrenting unlawful copies of the same works, all in violation of Plaintiffs’ copyrights.

The case is similar to Carreyrou in approach–“joined individual actions pursuant to Federal Rule of Civil Procedure 20”–but much larger, comprising 100 authors, authors’ estates, and publishers who either timely opted out or are seeking permission from the Bartz judge to opt out late.

Plaintiffs are seeking damages not just under the Copyright Act, but under 17 U.S.C. § 1203(c) of the Digital Millennium Copyright Act, which allows awards of up to $2,500 or up to $25,000, depending on the violation. They demand individual judgment on each of their claims, statutory damages “up to the maximum provided by law”, attorney fees and costs, a permanent injunction barring Anthropic and its agents from future infringement of “any of Plaintiffs’ exclusive rights under copyright”, and “[a]n order requiring that Defendants destroy under the Court’s supervision all infringing copies of Plaintiffs’ copyrighted works in Defendants’ possession or control.”

Again, this kind of joint action is “unorthodox”, in the words of some of the news coverage of it, and it remains to be seen how the court will respond–especially since the number of plaintiffs is so large and they are each demanding a separate jury trial.

Keeping Track

There’s a tremendous volume of AI litigation right now. Keeping track of it is a monumental task.

The best resource I’ve found for that is Chat GPT is Eating the World. This extensive resource provides overviews and updates for every case that’s being litigated (128 of them as of this writing), news and articles, a master case calendar, and more, including a data center lawsuit tracker. It’s an essential resource for keeping up to date with an issue of vital importance not just for writers, but for…everyone. (Forgive its clunky formatting.)

The post Anthropic Settlement Update: The Opt-Outs Strike Back appeared first on Writer Beware.

In December 2024, the FBI arrested the CEO, VP, and an associate of PageTurner Press and Media, one of the most prolific and predatory of the many Philippine publishing scams. The government determined that over the course of six years, PageTurner–which did business in the Philippines under the name Innocentrix–defrauded nearly 900 writers of more than $48 million.

The gears of US justice grind slowly (especially these days), but in early May, PageTurner CEO Michael Cris Traya Sordilla pleaded guilty to one count of wire fraud and one of money laundering. Although each count carries a maximum sentence of 20 years in prison, the government’s sentencing guidelines in Sordilla’s plea agreement suggest he will serve much less time.

The plea agreement also requires Sordilla to pay millions of dollars in restitution to his victims. Will anyone ever receive those payments, though? Today’s guest post by retired Professor of Economics Trent Bertrand takes a look at that question, using the PageTurner prosecution as a lens to examine how the prevalence of plea bargains in US criminal prosecutions often serves the needs of prosecutors and defense attorneys rather than the victims of all-too-common frauds. For Mr. Bertrand, the issue is not abstract, as you’ll see from his bio at the bottom of this post.

Can a Flawed US Legal System Discourage Fraud?

by Trent Bertrand

Scams have become so commonplace in the United States that the vast majority are never investigated or prosecuted. Though various administrations have highlighted the need to address widespread fraud, little attention has been given to a basic flaw in the US criminal legal system in either deterring fraud or bringing justice to its victims.

Most criminal prosecutions (some 96-98%) result in plea bargains rather than jury trials. It is not much of an exaggeration to suggest that the trial by jury criminal justice system has been replaced by a plea bargain system. The flaw in the system results from what the economics literature calls the ‘principal agent’ problem, where the incentives for the agents empowered to implement systems or policies do not align well with the principal goals of those systems or policies. The plea bargain system is better designed to meet the goals of the agents negotiating these pleas, i.e. the prosecutors and the defense attorneys, than it is in meeting the twin goals of deterring similar crimes and bringing justice to the victims.

The key feature of the system is a combination of sentencing guidelines that fall far short of the maximum sentences for the offenses being charged. This part of the plea bargain fully meets the needs of prosecutors to obtain guilty pleas without the risks and costs of trials, and for the defense attorneys to charge high fees for protecting their clients from the threat of those statutory maximum jail sentences. Plea bargains in fraud cases also include ‘orders for restitution’ for the victims, but with little incentive for the fraudsters to pay restitution, especially if the main culprits have laundered the money abroad and have returned to their home country.

In this article, I illustrate just how flawed the system is based on the likely outcome of the prosecution under way in San Diego of four defendants in the PageTurner Press and Media scam.

Arrests, a Guilty Plea, and a Problem of Enforcement

This scam took advantage of elderly authors anxious to get their works more widely recognized. The lead conspirator, Michael Cris Traya Sordilla, operated a call center from the Philippines to bilk some 800+ mainly senior authors of about $48 million with false claims about having their works published and marketed by reputable publishers or highlighted in films and telvision series.

Against the odds, three of the conspirators were arrested in San Diego in December of 2024, with a fourth indicted a year later, in December 2025. Three of them remain in custody.

On April 16, 2026, Sordilla signed a plea bargain arranged by the prosecutor Oleksandra Johnson and the Defense Attornery Antonio Yoon [USA vs.Michael Cris Traya Sordilla, Case No. 24CR2712-JLS, Plea Agreement, Filed May 6,2026, S.D. California]. The Sordilla plea bargain includes a recommended sentence based on the Federal guidelines for sentencing established for the two offenses, wire/mail fraud and money laundering. While the statutory maximum is 40 years in prison, the recommended sentence by the negotiating parties in the Sordilla plea is 33 months in prison. The actual sentence is at the sole discretion of the presiding judge but the record shows that the judges often reduce rather than enhance the recommended sentence.

Sordilla has been in custody since he was arrested in December of 2024. If, as is customary, he is given credit for time already served, he may have at most a few months more in jail if the plea bargain sentencing recommendations are accepted by the judge.

The government also confiscated over 6 million dollars held by the firms set up by the defendants as part of their scam. Sordilla ordered his coconspirators to launder some $42 million for deposit in bank accounts in the Philippines and, as part of the plea agreement will pay the $2.7 million that he claimed to have personally received as a fine.

What is in this plea bargain for the victims? Sordilla also agreed to an order of restitution of some $48.7 million equal to the amount known to have been fraudulently extracted from the victims. This may turn out to be just another gift to lawyers who may be hired by victims to try without much prospect of success to enforce the order of restitution with both the money and the culprits already outside the United States. The result of the plea bargain could well be a get out of jail card for a main conspirator soon to be living back in the Philippines with access to the bulk of the $42 million laundered abroad.

The DOJ’s announcement of the plea bargain [US Attorney’s Office, S.D. California, “Defendant Pleads Guilty in $48 Million Nationwide Book Publishing Scam Targeting Hundreds of Seniors”, May 7, 2026} puts a rosy spin on the outcome. US Attorney Adam Gordon is quoted as saying “The defendants didn’t just steal money- they stole dreams leaving victims with empty promises and devastating losses. Today’s guilty plea delivers justice for the victims…”. Special Agent in Charge Mark Remily of the FBI San Diego Field Office is quoted as saying “Today’s guilty plea is the beginning of holding all those involved in this Hollywood dream scheme accountable for their crimes. FBI San Diego, along with our law enforcement partners, remains steadfast in our pursuit of justice of any scammers attempting to steal Americans’ hard-earned money.” Would that it were true! A jail term of three years in prison and unenforceable order of restitution is a poor excuse for justice.

What Might Better Serve PageTurner (and Other) Fraud Victims?

Is there hope that common sense justice might be salvaged from this flawed system?

As with all federal plea bargains, the judge has the final say on sentencing and the defendant does not have a right to change his guilty plea. A solution would be to sentence the defendant to a significantly longer jail sentence, with parole or supervised release in line with the present guidelines contingent on meeting reasonable levels of restitution. There is scope for doing this under the present system, but the fly in the ointment is the short sentences prescribed under the guidelines. If Sordilla is facing just a few more months in prison, there is little incentive for him to scramble to fund the restitution fund in order to gain parole a few months ahead of his scheduled release.

The judge could set aside the plea agreement sentencing recommendation and impose a much longer sentence. In the PageTurner case, Sordilla is obligated under an addendum to the plea agreement to disclose all assets he owns or has transferred to other parties since the commencement of the scam. He is reportedly a very wealthy businessman in the Philippines, owner of several enterprises and well known for his philanthropic activities. A good starting point for putting common sense back into the system would be a lengthy sentence, say 30 years instead of the statutory maximum of 40 years, with parole provided in line with the present sentencing guidelines conditional on payment of a substantial restitution, perhaps $30 million rather than the unenforceable $48.7 million he agreed to in the plea bargain.

Such a sentence would provide the missing incentive for Sordilla to come up with the restitution funds, even if this implied the same sort of ‘devastating losses’ he imposed on his victims. This decision might be appealed by Sordilla, since the plea bargain gives Sordilla the right to appeal a sentence far out of line with the sentencing guidelines or to claim ineffective counsel, the two exceptions to the ‘no appeal’ rule included in the plea agreement. But appeals can fail, and such an action by the judge could bring justice to the victims.

Relying on a courageous judge to use his or her power over sentencing to create a significant incentive for the culprit to provide restitution funds may be wishful thinking. As emphasized in this analysis, the plea bargain system is designed to meet the needs of agents often more interested in racking up a conviction while minimizing demands on prosecutorial, court and prison resources than in providing justice to the victims. If justice is to be had in the PageTurner scam, the judge would have to craft the needed tools on his or her own and deviate from the Federal sentencing guidelines, albeit for a good reason. That is why the legal system for addressing fraud is inherently flawed. Unless the present system of sentencing guidelines is reformed to provide parole or supervised release only after the order of restitution is satisfied, the present legal system will continue to fail to provide deterrence to fraud or justice for its victims.

The lack of incentives for funding restitution by the guilty fraudsters is not the only flaw in the system. The legal literature also highlights the use of the system to entice innocent parties to plead guilty with greatly reduced sentences via plea bargains rather than to face the possibility of much harsher penalties at trial [Carissa Byrne Hessick’s book, Punishment Without Trial: Why Plea Bargaining is a Bad Deal provides an example]. This has some similar characteristics to the flaw analyzed in this article, in that this racks up recorded convictions for prosecutors and successful mitigation of potentially harsher sentences by defense attorneys without deterring crime or bringing justice to the victims.

What would a sensible reform of the dominant plea bargain justice system for fraud look like if incentives with the goals of a commonsense justice system were not so misaligned? A starting point would be to recognize the dominance of plea bargains and specifically define the sentencing guidelines as a recommendation for parole or supervised release conditional on meeting all the conditions of the plea bargain, including satisfaction of the order of restitution. Such an outcome would not be an outlier based on a single judge refusing to see deterrence of crime and justice for victims swept aside. The revised Federal Guidelines would instead create a strong incentive for the those admitting guilt to make restitution to their victims for their crimes.

Trent Bertrand is a retired Professor of Economics who has taught at the Johns Hopkins University, Cornell University, Queen’s University, McGill University, the State University of New York at Binghamton, Princeton University, Vanderbilt University, the University of Hawaii as well as participating in University Development Programs at Thammasat University in Thailand with Ford and Rockefeller Foundation support. A close friend and colleague of his at the SUNY Binghamton Department of Economics passed away with a broken spirit due to a belated realization that he had fallen victim to the PageTurner fraudsters.

The post Guest Blog Post: Can a Flawed US Legal System Discourage Fraud? A Look at the PageTurner Scam Prosecution appeared first on Writer Beware.

Recent developments in the world of copyright have been making many writers rethink their attitudes toward copyright registration and reversion of publishing rights.

Because many artificial intelligence companies used pirated books to train their large language models, there are now a growing number of copyright infringement class action lawsuits against them. While it is still undetermined whether these companies’ use of the copyrighted material was fair use or not, it has become clear that the use of copyrighted material from pirate libraries is a no-no, especially when the method involves torrenting, which means the companies participated in redistributing the materials.

The first of these lawsuits, Anthropic v. Bartz, just held the final Fairness Hearing on a class action settlement and, although there were some minor factors which delayed Judge Martinez-Olguin’s approval, it looks as if the class action settlement will be approved and 1.5 billion dollars will eventually be paid out to claimants who met the definition of the class.

Needless to say, this will be an unprecedented class action settlement involving copyright. As currently calculated, claimants for each copyrighted work that was pirated by Anthropic will share $3,100. If the work was self-published or the work’s rights had reverted to the author, they will receive the entire amount. It’s safe to say this is the first time the average writer will benefit from their copyright registration in any substantial way.

But not every writer benefited, for a number of reasons; the primary reason was that the book had to have had its copyright registered with the US Copyright Office. The definition of the class for the Anthropic class action was:

• have been downloaded by Anthropic from LibGen or PiLiMi;
• have an International Standard Book Number (ISBN) or Amazon Standard Identification Number (ASIN);
• have been registered with the United States Copyright Office within five years of the work’s first publication; and
• have been registered before being downloaded by Anthropic, or within three months of the work’s first publication.

Of the estimated seven million works that were pirated by Anthropic, less than 500,000 works were part of the class. As of the May 14 settlement hearing, the number of works claimed was 447,576. That’s about 7% of the pirated works.

The requirement that a work have an ISBN or ASIN is essentially unfair because it only recognizes individual books, but at least it doesn’t discriminate against self-published works. There is nothing in US Copyright Law that distinguishes between “books” and other literary works that may have a registered copyright. The requirement is only to make identifying works and verifying author and publisher easier for the settlement administrators. As I say, though, most books do have one or the other, even if the ASIN is connected to long out of print book being sold used. Presumably some book authors have managed to avoid Amazon entirely, but they must be a small number.

Another similar class action lawsuit, Elsevier Inc. v. Meta Platforms, Inc. was filed on May 5 by a bunch of publishers and Scott Turow as the only named author. It restricts the class even further. The proposed class definition is:

All legal or beneficial owners of registered copyrights, in whole or in part, for any book possessing an International Standard Book Number (ISBN) or journal article possessing a Digital Object Identifier (DOI) or International Standard Serial Number (ISSN), that Meta, without such owner’s authorization, (1) reproduced by downloading during torrenting and/or copying of web scrapes; or (2) distributed during torrenting; or (3) reproduced in connection with the development and/or training of a Llama Model. For purposes of this definition, copyrighted works are limited to those registered with the United States Copyright Office (a) within five years of the work’s publication and before being reproduced or distributed by Meta, or (b) within three months of publication.

The main difference from the Anthropic class is the limitation to only books that have ISBNs. ASINs don’t count, cutting out a large majority of self and indie published works, even if they do have registered copyrights. You can understand, I suppose, why the plaintiff publishers want the class restricted to the books that they published, but it’s even more grossly unfair to ebooks that were published without ISBNs because ISBNs are only important for physical book distribution. It’s hard to justify limiting a class action this way when, for all practical purposes, the fairer Anthropic settlement’s class definition worked (fingers crossed).

So will there now be a rush by indie authors to purchase ISBNs? It makes sense if, for example, you claimed a book without an ISBN in the Anthropic settlement, since there’s a good chance it will turn up again in the Turow class. Like with copyright registration and rights reversion, the effort and outlay start to look worthwhile. Large copyright class actions and settlements change everything.

Postscript. The fundamental problem is that there is no comprehensive registry for published works.

The post Has the Anthropic Settlement Changed Everything? appeared first on Writer Beware.

The deadline to file a claim in the $1.5 billion Bartz v. Anthropic copyright settlement passed at midnight on March 30, 2026. Now that all claims have been filed, I’m taking a look at where things stand and what’s yet to come.

If you need a refresh, my backgrounder on the settlement and the class action lawsuit that spawned it is here.

The Fairness Hearing is currently scheduled for May 14…

…though it could be further delayed. It’s already been pushed back once.

The hearing is a final step in the process of approving or rejecting the settlement. The court considers objections, opt-outs, attorneys’ fees, and the fairness of the settlement as a whole. Per the judge’s order, class members who’ve filed objections can speak at the hearing via Zoom (there’s a Zoom link in the order). I’ve heard from a couple of people who were hoping to attend in person; I don’t know what the status of that is.

Payouts to class members will be issued only after the judge has granted final approval, and any appeals of the settlement have been resolved. There’s currently no timeline for that; you can ignore the August 10 estimate in settlement FAQ #35, which was optimistic even last September when the FAQ was created. So don’t expect to get a check (or checks: payments may be issued in installments) anytime soon. On the plus side, the settlement fund, into which Anthropic is paying in installments, will be earning interest, so any delays will increase individual payments.

Disputes between rightsholders will be resolved by a court-appointed Special Master.

Class participation has been robust

The pirate datasets that Anthropic used for training comprised around seven million works. Because the settlement’s definition of the class excluded foreign works and works without a US copyright registration–with such works representing the vast majority of the pirated works–the estimated number of works eligible to be incuded in the settlement was only 500,000. The finalized list of eligible works wound up slightly smaller, at 482,460. (Settlement FAQs #51 and #52 explain how the list was compiled.)

According to the motion for final approval filed on March 19,

The Class’s participation in the Settlement has been outstanding. As of March 19, there are 99,450
claims for 264,809 Works, representing 54% of the Works List. By contrast, there are 350 opt-outs (less than 0.5% of the Works List) and 41 objections (only 32 of which are from Class Members).

If you’re wondering why participation of just over 54% qualifies as outstanding, participation in class action settlements is typically extremely low–below 10%, according to some estimates. Average payouts are correspondingly small. The Anthropic settlement, by contrast, guarantees a payout of at least $3,000 per eligible work–and unlike many other class actions, the settlement fund is non-reversionary, which means that any leftover funds will be distributed to class members, potentially making the payout even higher. Though many authors will have to split their payout with their publishers, the unusually large payout, along with major publicity about the settlement and extensive outreach by authors’ groups and others, has incentivized response.

The number of claims will certainly have risen in the two weeks between the filing of the motion and the March 30 deadline. Will they come close to the full number of eligible works? My bet is no, but stay tuned.

UPDATE 5/2/26: Well, this is why I don’t gamble. Per a status report filed April 16, class participation now stands at an extraordinary 91.3%. Claims have been filed for 440,490 of the 482,460 works on the Works List.

Only around 350 class members have opted out.

Lawyers’ fees have been reduced

Class counsel initially requested a fee of 20% of the settlement fund ($300 million plus several million in expenses), with a quarter of that amount going to the several additional law firms brought on to handle administration of the settlement. The original judge in the case (Judge Alsup, who retired at the end of December 2025), was extremely blunt in his opposition to this request, especially regarding payment to the proliferating law firms.

In a filing on March 20, class counsel capitulated, cutting their request nearly in half:

Plaintiffs move for (1) 12.5 percent of the $1.5+ billion Settlement Fund in attorneys’ fees
to Class Counsel; (2) reimbursement of $2,779,950.26 in litigation expenses incurred by Class
Counsel; (3) a cost reserve of $18,220,000 for future expenses, including payment to the Settlement
Administrator; and (4) service awards of $50,000 to each Class Representative.

Average lawyers’ fees in this type of class action typically range between 25%-35%.

Macmillan’s promise re: failure to register copyright

An unpleasant corollary of the Anthropic settlement has been writers’ discovery that their publishers may not have registered their copyrights as required by their contracts, making those works ineligible for the settlement even though they were part of the pirated datasets used by Anthropic. Alone among such publishers AFAIK, Macmillan has promised to address this lapse. From Publishers Lunch:

On that front, at least, there is some good news for Macmillan authors. The company shares with PL that it has been communicating to authors and agents who have inquired about unregistered copyrights, acknowledging, “From what we currently understand, this was largely our mistake and we take full responsibility. If your work was excluded from the settlement for this reason, we will make you whole by paying you what you otherwise would have been paid under the settlement.”

I’ve seen confirmation of this promise in a recent agency email.

It’s more complicated than just making payments for unregistered books, however. Macmillan will have to figure out whether the works were among the seven million included in the datasets. As far as I know, only one of the two datasets used by Anthropic (LibGen) is publicly available.

Anthropic wants to consolidate cases

Several authors who opted out of the settlement subsequently filed infringement lawsuits against Anthropic and several other AI companies. The lawsuits were filed as a group, but not as a class, with the plaintiffs seeking individualized awards of statutory damages to be determined by a jury.

In a March 25 filing, Anthropic gave notice that it has moved to sever these lawsuits’ claims against it from the claims against the other companies and, if severance is granted, that it will move to consolidate the claims with Bartz. While consolidation would presumably limit those authors’ restitution to the amount guaranteed by the settlement, which is considerably less than the up to $150,000 per work in statutory damages the authors are seeking, that’s less than pocket change to Anthropic, so I’d guess that it mostly doesn’t want to have to deal with the legal hassle of these individual suits. (The new judge in the case, Judge Martinez-Olguin, has indicated that she’s unlikely to grant Anthropic’s request.)

One objection to the settlement could make a significant difference…if the court agrees

In addition to claims and opt-outs, there have been objections to the settlement. As noted above, as of March 19 the number of objections stood at 41, 32 of which were from class members.

The motion summarizes these, starting on page 25, including those filed under seal (the court has ordered that these be unsealed, but as of this writing, most aren’t yet available on the docket). They include objections to the amount of the settlement (either that the total amount is inadequate or that the per-work award isn’t sufficient), objections to the amount of attorneys’ fees, objections to the court-approved notice sent to class members and/or the way it was distributed, objections from writers who missed the opt-out deadline and are seeking an exception, miscellaneous objections including one arguing that books that serve a “more important function” should recieve a higher payout than “fiction and humor” books (!), objections to the inclusion of publishers in the award distribution, and objections seeking to expand the class definition by including foreign and non-copyright registered works.

Regarding the latter two issues, one objection is especially interesting. Filed by law professor Lea Bishop, this objection takes issue not only with publishers’ inclusion in the distribution plan and the sharply-limited class definition that locks thousands of pirated works out of the settlement, but with class representation, alleging that class counsel and publishers coordinated to enable publishers to insert themselves as payees (in a post on the Authors Cuild’s discussion forum, the AG has disputed this characterization). The entire objection is worth reading, but here’s the basic argument:

Excluding publishers would obviously raise the payouts to authors. But including foreign and unregistered works would vastly expand the scope of the settlement, adding tens of thousands if not millions of works. If you’re thinking that would in turn reduce payouts, the terms of the settlement ensure otherwise (see page 13):

Finally, $1.5 billion (plus accrued interest) is the minimum size of the Settlement Fund….If the Works List ultimately exceeds 500,000 works, then Anthropic will pay an additional $3,000 per work that Anthropic adds to the Works List above 500,000 works.

This would be the exact nightmare Anthropic agreed to the settlement to prevent.

For that and other reasons, I’m thinking it’s a long shot. Ms. Bishop has asked to speak and present at the Fairness Hearing, but the judge has denied her request because she “is not a member of the class and therefore lacks standing to object to the settlement.”

Once again–stay tuned!

UPDATE 5/20/26: The Fairness Hearing was held on May 14 as scheduled, and by all accounts went smoothly. Several objectors were heard in person (multiple objections were also filed in writing), and it was noted that the opt-in rate stands at 92.77%–slightly higher than at the last update on April 30. This is extraordinary for a class action (of course, most don’t promise nearly the kind of payout included authors will receive from this settlement).

According to the Authors Alliance, most of Judge Martínez-Olguín’s questions “concerned attorneys’ fees and the structure of the cost reserve, not the merits of the objections.” It still looks like the per-work payout will be in the $3,000 range originally projected.

The judge wasn’t expected to approve the settlement at the hearing, and indeed she didn’t. Hopefully she won’t take too long, but she must consider objections to the settlement, and decide on the award for attorneys’ fees. Once approval is granted, there will likely be more waiting before any payouts are received: among other things, it’ll have to be determined how the payouts will be disbursed, and disputes between claimants will have to be resolved.

I’m following closely and will update as news comes in.

The post Anthropic Copyright Settlement: April Update appeared first on Writer Beware.