Connect with us

America

Project Panama: Inside Anthropic’s secret race to scan millions of physical books

Published

on

In early 2024, the artificial intelligence startup Anthropic initiated a clandestine operation dubbed “Project Panama,” an effort to “destructively scan” nearly every book in existence.

According to court filings obtained by The Washington Post, the company spent tens of millions of dollars over the course of a year to purchase millions of books and physically dismantle them by cutting off their spines. The pages were then scanned to feed vast quantities of information into the AI models powering products like the popular chatbot Claude.

The previously undisclosed details of Project Panama were revealed in more than 4,000 pages of documents related to a copyright lawsuit brought by authors against Anthropic, which investors recently valued at $183 billion. While the company agreed to a $1.5 billion settlement in August to resolve the case, a district judge’s decision last week to unseal a series of related documents has provided a clearer picture of Anthropic’s aggressive pursuit of literary data.

These new filings, alongside documents submitted in other copyright cases against AI firms, illustrate the extraordinary lengths to which technology giants—including Anthropic, Meta, Google, and OpenAI—have gone to acquire massive troves of data for “training” their software. The Anthropic litigation is part of a broader wave of legal action by authors, artists, photographers, and news organizations who claim their creative works are being exploited.

Court records reveal that these companies view books as a premier prize. In a January 2023 document, one of Anthropic’s co-founders suggested that training AI models on books would teach them “to write well,” rather than merely mimicking “low-quality internet slang.” Similarly, a 2024 internal Meta email described access to digital book archives as “essential” for staying competitive with AI rivals.

However, the records also show that these companies found it “impractical” to obtain direct permission from publishers and authors. Instead, Anthropic, Meta, and others devised ways to acquire books in bulk without the authors’ knowledge. According to court records, these methods included downloading pirated copies.

When Anthropic launched Project Panama to purchase and scan physical books, it turned to a Silicon Valley veteran. The company hired Tom Turvey, a former Google executive who two decades ago helped spearhead the famous but legally controversial Google Books project.

Anthropic initially considered sourcing books from libraries or iconic second-hand bookstores like New York City’s Strand, famous for its “18 miles” of new and used titles. A March 2024 document detailing an Anthropic content acquisition meeting noted that the store “was interested in providing second-hand books.” Documents also show Anthropic employees discussed approaching US libraries, including the New York Public Library or even “a chronically underfunded new library.”

It remains unclear which of these proposals, if any, were executed. A spokesperson for Strand, reached via email, stated that the bookstore did not sell any books to Anthropic.

Ultimately, documents indicate that Anthropic purchased millions of books, often in batches of tens of thousands, relying on used-book retailers such as Better World Books and the UK-based World of Books. While the final number of scanned books and the total cost were redacted in the documents, a project proposal from a vendor working with Anthropic specified that the AI firm was seeking a “document scanning service provider experienced in converting 500,000 to two million books over a six-month period.”

The document explained that the scanning firm’s “hydraulic-driven cutting machine” would “neatly cut” the books, and the pages would then be scanned using “high-speed, high-quality, production-level scanners.” Finally, the vendor would arrange a schedule with a “recycling company to collect the completed books.”

Internal messages show that Meta employees repeatedly expressed concerns that downloading millions of books without permission would violate copyright law. In December 2023, an internal email submitted in the copyright case against Meta noted that the practice was approved after being “communicated to MZ,” an apparent reference to CEO Mark Zuckerberg.

In a recently released legal filing, Anthropic revealed that co-founder Ben Mann personally spent 11 days in June 2021 downloading fiction and non-fiction titles from “LibGen,” a well-known “shadow library” hosting pirated books and other copyrighted content. A screenshot of a web browser included in the files showed Mann using file-sharing software to download the data.

A year later, in July 2022, Mann welcomed the launch of a new website called Pirate Library Mirror, which claimed to host a massive database of books and stated, “we are intentionally violating copyright law in most countries.” Mann sent a link to the site to other Anthropic employees with the message: “just in time!!!”

In legal filings, Anthropic argued that it did not train a commercial AI model for profit using LibGen data and that it never used Pirate Library Mirror to train any completed AI model.

Ed Newton-Rex, a former AI executive and music composer who now leads a non-profit advocating for creators’ rights, said these revelations underscore that AI companies owe creators far more than they have paid to date. “We urgently need a reset in the AI industry so that creators start getting paid fairly for the vital contributions they make,” he said.

Google, Microsoft, and ChatGPT-maker OpenAI face similar copyright lawsuits from authors. While many of these cases remain pending, James Grimmelmann, a professor of digital and information law at Cornell Tech, noted that the legal questions they raise remain unresolved.

However, in two separate rulings, judges determined that tech companies’ use of books to train AI models without author or publisher permission might be legal under the “fair use” doctrine of copyright law. In June, District Judge William Alsup ruled that Anthropic had the right to use books for training because they processed the works in a “transformative” manner. The judge likened the AI training process to teachers “teaching school children how to write well.”

That same month, District Judge Vince Chhabria ruled in the Meta case that authors failed to prove the company’s AI models could harm the sales of their books.

Nevertheless, companies may still face legal jeopardy regarding how they acquired the books. In the Anthropic case, while the scanning project was accepted, the judge ruled that the company may have violated authors’ copyrights by downloading millions of pirated books for free before launching Project Panama. Alsup granted class-action status to authors whose works were included in two shadow libraries that Anthropic downloaded and stored for future use.

Rather than go to trial, Anthropic agreed to pay publishers and authors $1.5 billion without admitting wrongdoing. Authors whose books were downloaded can claim a share of the settlement, estimated at approximately $3,000 per book.

Aparna Sridhar, Anthropic’s deputy general counsel, stated in an email to The Washington Post: “This case has been resolved, but the court’s landmark June 2025 ruling remains valid. Judge Alsup argued that AI training is ‘fundamentally transformative’: Anthropic’s AI models were trained ‘not to copy or replace works, but to get over a difficult hump and create something different.’ What we settled on was how some materials were obtained, not whether we could use them to develop AI models.”

Documents released in the Meta lawsuit suggest that the social media giant’s employees were equally data-hungry and willing to take legal risks to obtain it. While Judge Chhabria sided with Meta on the use of books for training, he allowed authors to proceed with claims that Meta illegally distributed copies of pirated books. The plaintiffs are seeking class-action status for these claims in the Northern District of California.

In that case, authors alleged that Meta’s senior executives considered purchasing books for training but instead opted to download millions of books for free from “torrent” platforms that facilitate online piracy. Internal documents, some previously reported, show Meta employees expressing concerns that their actions were risky or wrong and discussing how to hide their tracks.

One engineer wrote in 2023, “Downloading torrents from a company laptop doesn’t feel right.” The same employee later voiced concern to the legal team that using torrent sites might require sharing pirated works with others, which “might not be legally appropriate.”

A December 2023 email clearly stated that the use of LibGen was approved after Zuckerberg was notified. “After prior notification to MZ, GenAI’s use of LibGen for Llama 3 was approved… with a series of agreed-upon mitigating measures,” the email read, before listing legal and political risks. It noted that media reports suggesting the use of a known pirate dataset like LibGen could “weaken our negotiating position with regulators on these issues.”

By April 2024, internal correspondence showed the company moving to download LibGen and other shadow libraries. Chat logs show one employee asking another why they were using servers rented from Amazon for torrenting instead of Facebook-owned servers. The answer: “to avoid the risk of the activity being traced back to the company.”

In a filing last month, Meta’s lawyers wrote that the company “denies distributing the plaintiffs’ works while downloading training data using torrents.”

In a separate 2023 case, authors accused OpenAI and Microsoft of violating copyright law by using books for AI training. OpenAI, where Mann and Anthropic CEO Dario Amodei worked before founding their own firm, admitted to downloading LibGen but told the court it deleted the files before the launch of ChatGPT.

Justin A. Nelson, an attorney at Susman Godfrey LLP representing authors in both the OpenAI and Anthropic cases, said: “OpenAI fired the opening shot that led to the widespread piracy by AI companies and the exploitation of all human expression.”

Earlier this month, two major publishers applied to join a group of authors and illustrators in a 2023 copyright lawsuit against Google.

Grimmelmann, the Cornell Tech law professor, observed that AI companies “led themselves into a delusion” regarding the use of copyrighted data. The breakthroughs behind tools like ChatGPT began in academic research, where the use of copyrighted material for training is widely accepted, but researchers continued the practice even as AI models became commercialized.

“By the time the tension became apparent, they had invested heavily in incorporating copyrighted data into their workflows and were in a fast-paced, high-stakes competition to launch newer and better models,” Grimmelmann said.

America

Iran team leaves thank-you message in Los Angeles locker room after World Cup draw

Published

on

Iran’s national football team left a message in its locker room at SoFi Stadium, thanking Los Angeles for its hospitality during the World Cup.

The players said they were leaving the city with honor after keeping their hopes of reaching the knockout stage alive with a 0-0 draw against Belgium.

In the handwritten note, published by the Iran Football Federation, the team wrote:

“From the ancient land of Persia thousands of years ago to the civilized Iran of today, the spirit of Iran remains alive and unshaken. Los Angeles, thank you for your hospitality. We arrived in Los Angeles with pride, competed with honor and leave with dignity.”

The note also thanked Iranian supporters who gave their “hearts, voices and souls” to the team throughout its two matches and concluded with a call for peace, respect and friendship among all nations.

Los Angeles hosted both of Iran’s Group G matches, while the team returned to its training base in Tijuana between games.

Iran has been based in Tijuana throughout the tournament and has had to travel back and forth to the United States for matches because of restrictions related to its stay in the country. Entry bans were also imposed on some members of the national team’s coaching staff and officials.

US authorities said the team’s travel arrangements remain under review, while discussions continue over the possible easing of some restrictions.

Iran head coach Emir Ghalenoei has repeatedly criticized the travel restrictions, saying his squad has faced challenges that no other team in the tournament has been required to endure.

After drawing 2-2 with New Zealand in its opening match at SoFi Stadium, Iran will play its final Group G match against Egypt in Seattle.

Continue Reading

America

Colombia’s de la Espriella claims narrow presidential victory in runoff election

Published

on

The first results from Colombia’s presidential runoff election showed that right-wing candidate Abelardo de la Espriella, backed by Donald Trump, had narrowly won the vote.

The victory of de la Espriella, who has no prior political experience, signals a fundamental shift in the government’s approach to tackling the country’s long-running internal armed conflict and rising violence.

Throughout the campaign, de la Espriella pledged to intensify military pressure on illegal armed groups, drug trafficking networks and criminal organizations. He succeeded in defeating left-wing candidate Iván Cepeda, a close ally of incumbent President Gustavo Petro.

Speaking after the initial results were released, de la Espriella said: “Today marks the beginning of a new era for our country. This era is built on the free and democratic will of millions of citizens who chose to believe in a great, secure, prosperous Colombia full of opportunities.”

Cepeda says he will await official results

According to the preliminary count, with more than 99% of ballots tallied in the runoff election, de la Espriella secured approximately 49.7% of the vote, while Cepeda received 48.7%.

Cepeda, who has not yet conceded defeat, said the preliminary results were neither official nor binding.

“When the official count is completed, the final results are known and the necessary verification procedures are finished, we will recognize the official outcome produced by that process,” Cepeda said.

Reuters reported that the verification process showed very little variation from the preliminary counts recorded during the first round of voting on May 31.

De la Espriella, who grew up in Colombia’s Caribbean region, drew particularly strong support from that part of the country. Addressing a large crowd gathered in the coastal city of Barranquilla after the first results emerged, de la Espriella, who has adopted the nickname “El Tigre” (The Tiger), declared: “Tonight is the beginning of a new story for the nation. Tonight a new era begins, a change of order begins.”

He said he would govern for all Colombians, including those who voted for his opponent, and pledged loyalty to and protection of Colombia’s 1991 constitution.

At celebrations in Barranquilla, supporters wore Colombia’s yellow national football jersey and waved Colombian flags.

With images of de la Espriella projected behind the stage, supporters chanted “Stand firm for the homeland” and “Petro out!” as fireworks lit the sky. Some supporters wore hats bearing the slogan “Make Colombia Great Again,” echoing those worn by supporters of US President Donald Trump.

Trump reacted to the results in a Truth Social post, writing: “BIG won!”

One supporter, Patricia, told reporters: “We are tired of the murders in this country and of this government’s bureaucracy. Now we finally have a president from the coastal region.”

Another supporter said: “We are proud of the Tiger. We hope he transforms the country and, above all, creates a new nation where we will have jobs and greater security.”

Supporters of Cepeda, who narrowly lost the election, also voiced concerns on the streets of Barranquilla.

Catalina La Grande, a student and activist who supports Cepeda, told the BBC: “There is a visible sense of unease in the air. Such a narrow margin worries us because it reflects how divided the country is and the enormous challenges we face in defending democracy, peace and human rights.”

Another young voter backing Cepeda, Maria, said the results showed a divided country but noted that the public had remained peaceful.

“Given the level of polarization we are experiencing, the absence of violence in the streets is a positive development,” she said.

The sharp divisions between the candidates have fueled concerns that unrest could emerge if some opposition groups refuse to accept the outcome.

Late on Sunday night, clashes were reported between protesters and police in Cali, Colombia’s third-largest city. Demonstrators reportedly burned US flags, while police used tear gas to disperse large crowds angered by de la Espriella’s victory.

President Gustavo Petro is also reported to be considering challenging the result. In a post on X, Petro said that based on the preliminary count, “no one can be declared president” and alleged that the security of some polling stations had been compromised. He called for an audit of the voting software but provided no evidence to support the claims.

Who is Abelardo de la Espriella?

De la Espriella, who has no political background, is a lawyer and businessman. During his legal career, he represented clients including Alex Saab, an ally of former Venezuelan President Nicolás Maduro who has faced money laundering charges in the US, and David Murcia Guzman, one of Colombia’s most notorious fraudsters.

De la Espriella says he handled those cases in his capacity as a defense attorney.

Often compared to El Salvador’s President Nayib Bukele because of his security policies and distinctive beard, de la Espriella and his supporters frequently wear Colombia’s national football jersey at rallies and on social media. Critics accuse him of politicizing the national team shirt.

He is also known for regularly addressing campaign crowds from behind bulletproof glass panels.

Colombia’s internal armed conflict has persisted for decades, but violence has intensified in recent years. Armed groups and criminal organizations, including dissident factions of the FARC, the ELN and the Clan del Golfo, have doubled their membership over the past five years.

Competition for control of lucrative cocaine trafficking routes and illegal mining operations has further escalated the violence. Fighting along the Colombia-Venezuela border last year displaced tens of thousands of people. Cocaine production in the world’s largest cocaine-producing country has reached record levels.

Critics of President Petro argue that his “total peace” strategy, which prioritizes negotiations with armed groups, has failed, claiming that such groups have used ceasefire arrangements to expand their territorial control and influence.

De la Espriella has pledged to cancel all negotiations with illegal armed groups and increase military pressure to restore order.

As part of that agenda, he has promised closer cooperation with the US, the construction of massive prisons in Colombia’s forests, a smaller state apparatus and reforms to the healthcare system.

Having lived and worked in Miami for many years, de la Espriella has held US citizenship since 2023. During the election campaign, he received support from Donald Trump, who said de la Espriella would “stop illegal migration, fight crime and drugs, and restore law and order.”

Before the election, Trump also said de la Espriella would feel “the full support and strength of the United States” behind him.

Although Colombia has historically been one of Washington’s closest allies in the region, relations have become strained in recent years due to sharp disagreements between President Trump and President Petro over migration policy, tariffs and military intervention in Latin America.

De la Espriella’s election also aligns with a broader trend across Latin America, where security concerns have pushed politics to the right. His victory was welcomed by other conservative leaders across the region.

Argentine President Javier Milei said Colombians had “chosen the path of economic freedom, prosperity and uncompromising security” and had declared that enough was enough to transnational organized crime and drug trafficking.

Chile’s José Antonio Kast said: “A new era of freedom is beginning for Colombia, one that will allow the country to regain security and prosperity.”

Continue Reading

America

Data leak exposes Peter Thiel’s secret ‘Dialog’ network of politicians, regulators, and tech elites

Published

on

“Dialog,” an exclusive, highly secretive association founded by billionaire technology investor and Palantir co-founder Peter Thiel, has had its internal records exposed online. The leaked documents reveal the identities of influential figures spanning US politics, finance, national security, and technology who have participated in the group’s private gatherings.

According to a report by WIRED, the society’s internal directories and registry records were left exposed on the internet, laying bare the attendee lists for its highly restricted events.

Established by Thiel in 2006, Dialog is an invitation-only private organization that convenes US officials, foreign government representatives, and Silicon Valley executives for off-the-record annual retreats. For nearly two decades, the organization has consistently declined to disclose its membership rolls.

The exposure began when a directory embedded in the source code of the group’s website was discovered by Swiss hacker maia arson crimew. Known for previously exposing the US government’s No Fly List and breaching the systems of security camera firm Verkada, the hacker told WIRED that the directory was uncovered via an anonymous tip. WIRED independently verified the authenticity of the directory’s contents.

Separately, a source provided WIRED with the registration list for Dialog’s upcoming 2026 retreat.

The list features the names of 222 individuals, detailing each attendee’s membership status and designation, including classifications such as “active member” and “guest.” The retreat is scheduled to take place from August 12 to 16, 2026, at the Powerscourt Hotel, located just outside Dublin, Ireland.

The leaked data also details a program composed of closed-door sessions. Panel titles include: “Does Money (Really?) Buy Happiness,” “Bring Back Nuclear Power,” “Navigating WWIII,” “Battlefield Technologies,” and “How Is Your Sex Life?”

Other scheduled discussions include a session titled “Start a Cult,” led by the founder of the Christian networking site Pray.com, and “Start a Party,” conducted by a former White House national security official.

While some of the documents contain the standard material of typical thought-leadership conferences, they also reveal an extraordinary convergence of institutional and private power.

Among the registry records is General Alexus Grynkewich, the Commander of US Air Forces Central Command who took office as the Supreme Allied Commander Europe of NATO and Commander of US European Command in July 2025. The leaked list indicates that Grynkewich has attended Dialog events since 2021.

The website directory also lists current Trump administration officials, two US senators, six members of the “PayPal Mafia,” a former Middle Eastern intelligence chief, and a sitting US ambassador. These names appear alongside the founders and executives of some of the United States’ largest surveillance, data brokerage, and advertising data firms.

These industry executives appear in the directory alongside the very regulators and lawmakers who oversee their sectors.

For instance, Dialog’s chairman, Auren Hoffman, co-founded SafeGraph, a location data broker, and LiveRamp, an identity-resolution firm—two pivotal suppliers in the consumer data economy.

Hoffman is listed in the directory alongside Treasury Secretary Scott Bessent, who shapes financial data regulations, and Senator Ted Cruz, the Chairman of the Committee on Commerce, Science, and Transportation, which oversees the Federal Trade Commission and data privacy authorities.

Similarly, Palantir co-founder Joe Lonsdale is listed in the same association directory as Secretary of the Army Dan Driscoll and Representative Jim Himes, the ranking member of the House Intelligence Committee, which oversees the intelligence agencies with which Palantir contracts.

None of the individuals named in the WIRED report responded to requests for comment. Raffi Grinberg, who identifies himself on LinkedIn as Dialog’s executive director and is the author of the self-help book How to Be a Grown-Up, also did not respond to inquiries.

The registry records appear to document not only who holds membership in Dialog, but also who has confirmed attendance for upcoming events.

Of the 222 individuals registered for the 2026 event, 87 are marked as first-time attendees, according to the leaked records.

Others have histories with the group spanning more than a decade, with a select few dating back to the society’s founding 20 years ago. None of the registrants, including Grynkewich, utilized official government email addresses to register.

Instead, all participants registered using personal or corporate email accounts, effectively keeping their correspondence and attendance outside the purview of public records laws that apply to government email systems.

According to WIRED, the primary thread uniting the attendee list—beyond titles and offices—is a shared interest in artificial intelligence, longevity, and the near future.

When prompted on the registration form to predict future trends, registrants repeatedly returned to a central theme: that artificial intelligence will thoroughly restructure commerce, warfare, education, and faith within a few years.

Several participants predicted mass unemployment and a subsequent return to labor unions and state-sponsored programs. Others anticipated an “AI winter,” domestic terrorism targeting data centers, defendants choosing AI-powered attorneys over public defenders, or a religious revival triggered by these technological shifts.

Another registrant predicted that “societal decay will continue to accelerate.”

Members also listed personal skills on their forms, ranging from “funhouse construction,” accent mimicry, backcountry skiing, and urban exploration to “meditative and psychedelic exploration into the nature of reality.”

One participant listed “compassion and existential dread” as personal talents, while another listed “dinner parties, keeping secrets, remembering birthdays.”

The book recommendations listed by participants are classic and optimization-oriented. Recommended texts include works by Marcus Aurelius and Milan Kundera, alongside Annie Duke’s Thinking in Bets, Peter Attia’s Outlive, and—recommended by at least one attendee—Thiel’s own book, Zero to One.

Dialog also facilitates personal matchmaking. The registration form asks participants, “Are you looking for love?” and offers options to opt into a “future matchmaking” database as a “Single Male,” “Single Female,” or “Other.”

A separate website, dating.dialog.org, hosts an application promoted with the slogan: “meaningful connections for extraordinary people.”

The form also collects sensitive personal data, including the “political leaning” of each participant. Dialog pledges on the form that this information “will never be shared on the app or with other attendees.” However, this data and the matchmaking responses were both compromised in the leak.

The records are maintained on Airtable, a commercial database platform. For each participant, Dialog records their membership status, a history of all retreats attended, a biography, their city of residence, and a personal access token.

The leaked registry contains high-profile names that do not appear in the group’s public-facing directory. Among them are Randy Kroszner, a former Federal Reserve Governor who currently serves on the Bank of England’s Financial Policy Committee; Hallie Hoffman, former general counsel and acting chief of staff for the Drug Enforcement Administration (DEA); Jonathan Greenblatt, CEO of the Anti-Defamation League (ADL); Peter Goettler, President of the Cato Institute; Ryan Stowers, executive director of the Charles Koch Foundation; and Roger Myerson, a Nobel laureate economist from the University of Chicago.

The list also includes a contingent of Google and Google DeepMind executives, including Tom Lue, who manages global relations for the company’s pioneer AI division, as well as Washington Post national security correspondent Souad Mekhennet.

The remainder of the roster consists of hedge fund and private equity billionaires, current and former foreign officials, television actors, bestselling authors, and religious leaders.

Among the internal documents left exposed on the database was a guide for event moderators.

The document instructs moderators to remind participants that all discussions are strictly “off-the-record” and that comments must be concise and “non-obvious.”

Moderators are also instructed to keep introductions brief to “avoid status signifiers” in rooms crowded with senators, high-ranking officials, and corporate leaders.

Since its inception, Dialog has operated with a minimal public footprint. It hosts at least one meeting per year where seating is pre-assigned, sessions are structured by moderators, and strict non-attribution rules apply to all remarks.

According to Axios, which first reported on the group’s plans to establish a physical campus in the Washington, DC area, past meetings have been held at the Ritz-Carlton Dove Mountain in Arizona and the San Clemente Palace in Venice, Italy.

The society has frequently been described as a technology-sector equivalent of the Bilderberg meetings, where Western political and business elites convene behind closed doors.

In 2022, statistician Andrew Gelman published one of Dialog’s invitation letters on his blog, revealing the event’s format and a registration fee exceeding $16,000.

The invitation list for the group’s 2014 meeting re-entered public scrutiny this year after the US Department of Justice released files showing Jeffrey Epstein was among approximately 150 invitees. Whether Epstein actually attended the meeting remains unclear.

Continue Reading

MOST READ

Turkey