Anthropic has agreed to pay at least $1.5 billion to settle a lawsuit brought by a group of book authors alleging copyright infringement, an estimated $3,000 per work. In a court motion on Friday, the plaintiffs emphasized that the terms of the settlement are “critical victories” and that going to trial would have been an “enormous” risk.

This is the first class action settlement centered on AI and copyright in the United States, and the outcome may shape how regulators and creative industries approach the legal debate over generative AI and intellectual property. According to the settlement agreement, the class action will apply to approximately 500,000 works, but that number may go up once the list of pirated materials is finalized. For every additional work, the artificial intelligence company will pay an extra $3,000. Plaintiffs plan to deliver a final list of works to the court by October.

“This landmark settlement far surpasses any other known copyright recovery. It is the first of its kind in the AI era. It will provide meaningful compensation for each class work and sets a precedent requiring AI companies to pay copyright owners. This settlement sends a powerful message to AI companies and creators alike that taking copyrighted works from these pirate websites is wrong,” says colead plaintiffs’ counsel Justin Nelson of Susman Godfrey LLP.

Anthropic is not admitting any wrongdoing or liability. “Today’s settlement, if approved, will resolve the plaintiffs’ remaining legacy claims. We remain committed to developing safe AI systems that help people and organizations extend their capabilities, advance scientific discovery, and solve complex problems,” Anthropic deputy general counsel Aparna Sridhar said in a statement.

The lawsuit, which was originally filed in 2024 in the US District Court for the Northern District of California, was part of a larger ongoing wave of copyright litigation brought against tech companies over the data they used to train artificial intelligence programs. Authors Andrea Bartz, Kirk Wallace Johnson, and Charles Graeber alleged that Anthropic trained its large language models on their work without permission, violating copyright law.

This June, senior district judge William Alsup ruled that Anthropic’s AI training was shielded by the “fair use” doctrine, which allows unauthorized use of copyrighted works under certain conditions. It was a win for the tech company but came with a major caveat. As it gathered materials to train its AI tools, Anthropic had relied on a corpus of books pirated from so-called “shadow libraries,” including the notorious site LibGen, and Alsup determined that the authors should still be able to bring Anthropic to trial in a class action over pirating their work. (Anthropic maintains that it did not actually train its products on the pirated works, instead opting to purchase copies of books.)

“Anthropic downloaded over seven million pirated copies of books, paid nothing, and kept these pirated copies in its library even after deciding it would not use them to train its AI (at all or ever again). Authors argue Anthropic should have paid for these pirated library copies. This order agrees,” Alsup wrote in his summary judgement.

Researchers at Institute of Science Tokyo have created a new material platform for non-volatile memories using covalent organic frameworks (COFs), which are crystalline solids with high thermal stability. The researchers successfully installed electric-field-responsive dipolar rotors into COFs.

Due to the unique structure of the COFs, the dipolar rotors can flip in response to an electric field without being hampered by a steric hindrance from the surroundings, and their orientation can be held at ambient temperature for a long time, which are necessary conditions for non-volatile memories. The study is published in the Journal of the American Chemical Society.

Humans have made great efforts to record information by inventing recording media such as clay, paper, compact disks, and semiconductor memories. As the physical entity that holds information—such as indentations, characters, pits, or transistors—becomes smaller and its areal density becomes higher, the information is stored with higher density. In rewritable memories, the class called “non-volatile memories” are suitable for storing data for a long time, such as for days and years.

Recently, molecular technology has evolved. One class of molecular technology consists of molecules that exhibit mechanical motions. They are called “molecular machines” or “nanomachines.” If a mechanical entity rotates or flips around a chemical bond, which serves as an axis, the material class is particularly called “molecular rotors.”

Use of molecular rotors to store information may cause a breakthrough. This is because the size of molecules is a few orders of magnitude smaller than the sizes of pits in a compact disk and transistors in semiconductor memories, and organic molecules are inherently highly designable. Although applications using molecular machines have been explored extensively, the attempts to develop non-volatile memories have been scarce, mainly because the simultaneous satisfaction of the following three requisites has been so challenging.

1. To control the orientation of molecular rotors with an electric field, the rotors have to have a dipole—a spatial displacement of a positive charge and a negative charge necessary to gain a force from the applied electric field.
2. The rotors must not flip at ambient temperatures so that their orientations are held for a long period.
3. There must be adequate spaces around the rotors so that they can flip without being hampered by the steric hindrance that may be caused by the tight packing of the molecules in the solid phase. Additionally, the substance has to be heat durable up to the temperatures current computational components ordinarily undergo, which is often up to 150°C.

New materials developed by the researchers of Institute of Science Tokyo have achieved these three requisites simultaneously, with very high thermal durability up to near 400°C. By demonstrating these novelties for the first time, the researchers have created a material foundation for molecular-machine-based non-volatile memories that potentially store information at higher density than current technologies.

The researchers selected covalent organic frameworks (COFs) as a platform for the aim. COFs are an emerging class of crystalline solids formed by periodically connecting two kinds of building block molecules by covalent bonds. For one building block, they chose a tetrahedral, four-handed molecule. For the other building block, they newly developed a flat, three-handed molecule in which three dipolar rotors (1,2-difluorophenyl, DFP) and three aryl groups are alternately positioned around the central benzene ring.

Previously, these aryl groups were shown to suppress the flip of the DFP rotors at ambient temperatures in a toluene solution, which satisfied requisites 1 and 2 above, but the high density of the molecular solid sterically hindered the flip of the rotors in the solid phase, which could not satisfy requisite 3.

Interestingly, the COFs they developed exhibited an unprecedented shape dimorphism, in which the COFs grew to a hexagonal prism shape or a membrane shape, depending on the solvent composition used for the growth. Furthermore, from X-ray structural analyses, these new COFs turned out to have an unprecedented sln topology, which has a low density inherently and has not been reported for COFs.

“Due to the substantially low density of about 0.2 g/cm³ caused by the unique sln topology possessed by the COFs, the dipole rotors incorporated into the periodic network constituting the COFs have adequate spaces around them, allowing them to flip without suffering from the steric hindrance from their surroundings.

“This is a breakthrough, because our COFs are a rare solid in which dipolar rotors can flip when they are brought to elevated temperatures above 200°C or undergo sufficiently strong electric fields, but their orientations can be held for a long time at ambient temperatures. These uniquenesses have been realized by our careful selection of the building block molecules to create the COFs for this aim,” says Professor Yoichi Murakami, the leader of this project.

Additionally, Murakami pointed out the significance of the work also exists in the extension of the diversity of COFs by their discoveries of sln topology and shape dimorphism, both of which were unknown for COFs previously.

These COF-based solids may be a new platform for storing information with further higher density after proper scale-up and device demonstration are made subsequently.

Setting up and migrating to Dashlane from another password manager is simple, and you’ll use a secret key to encrypt your passwords, much like BitWarden’s setup process. In practice, Dashlane is very similar to the others on this list. Dashlane offers a 30-day free trial, so you can test it out before committing.

After signing up, download the app for Android and iOS, and grab the browser extensions for Firefox, Chrome, and Edge.

Best for Bundled Services

You might know Nord better for its VPN service, but the company also offers a password manager, NordPass, and a pretty nice online storage system, NordLocker. A part of the appeal of NordPass comes in bundling it with the company’s other services for some compelling deals. As a password manager, NordPass offers everything you need. It uses a zero-knowledge setup in which all data is encrypted on your device before it’s uploaded to the company’s servers. Unlike most services here, NordPass uses XChaCha20 for encryption. It would require a deep dive into cryptography to get into the differences, but the short story is that it’s just as secure and maybe slightly faster than the AES-256 encryption used by other services.

There’s a personal information storage feature to keep your address, phone number, and other personal data safe and secure, but easy to access. NordPass also offers an emergency access feature, which allows you to grant another NordPass user emergency access to your vault. It works just like the same feature in 1Password, allowing trusted friends or family to access your account if you cannot.

Other nice features include support for two-factor authentication to sign in to your account, as well as security tools to evaluate the strength of your passwords and alert you if any of your data is compromised. Note that NordPass Premium is theoretically $3 a month, but there are always sales that bring that much lower.

The downside, and my one gripe about all Nord services, is that there is no monthly plan. As noted above, the best deal comes in combining NordPass, NordVPN, and NordLocker for a bundled deal. A free version of NordPass is available, but it’s restricted to only a single device.

After signing up, download the app for Android and iOS, and grab the browser extensions for Firefox, Chrome, and Edge.

Best DIY Options (Self-Hosted)

Want to retain more control over your data in the cloud? Sync your password vault yourself. The services below do not store any of your data on their servers. This means attackers have nothing to target. Instead of storing your passwords, these services use a local vault to store your data, and then you can sync that vault using a file-syncing service like Dropbox, NextCloud, or Edward Snowden’s recommended service, SpiderOak. There are two services to keep track of in this scenario, making it a little more complex. But if you’re already using a file-syncing file service, this can be a good option.

You can also properly host your own vault with network-attached storage or a local server.

Courtesy of Enpass

Enpass does not store any data on its servers. Syncing is handled through third-party services. Enpass doesn’t do the syncing, but it does offer apps on every platform. That means once you have syncing set up, it works just like any other service. And you don’t have to worry about Enpass being hacked, because your data isn’t on its servers. Enpass supports syncing through Dropbox, Google Drive, OneDrive, iCloud, Box, Nextcloud, or any service using WebDAV. Alas, SpiderOak is not currently supported. You can also synchronize your data over a local WLAN or Wi-Fi network.

All of the features you expect in a password manager are here, including auto-generating passwords, breach-monitoring, biometric login (for devices that support it), auto-filling passwords, and options to store other types of data, like credit cards and identification data. There’s also a password audit feature to highlight any weak or duplicate passwords in your vault. One extra I particularly like is the ability to tag passwords for easier searching. Enpass also makes setting up the syncing through the service of your choice very easy. Enpass added support for passkeys, too.