Imagine if McDonald’s could use trademark law to control how you use the term “fast food.” Or if the Canadian government could stop you from using the word “Canada” in the title of a book about the country and its people. That wouldn’t just be absurd; it would be an unacceptable obstacle to criticism of and commentary about those institutions. Yet the Church of Jesus Christ of Latter-day Saints (the “LDS Church”) has a track record of claiming exactly that kind of authority over the word “Mormon,” using the threat of expensive litigation to pressure speakers into compliance.

We at EFF have opposed the LDS Church’s abuse of trademark law for over a decade. In 2014, we filed an amicus brief when the church sued an online dating service for church members called Mormon Match. In 2016, it threatened legal action against our client the Mormon Mental Health Association, a nonprofit association for mental health professionals who work with members of Mormon faiths. In 2025, the church tried to pressure our client Burke Sorenson into changing the name of his Mormon News Roundup podcast. Now, the LDS Church has brought a lawsuit over a podcast called Mormon Stories that examines Mormonism and Mormon culture. With the help of attorneys at Ballard Spahr, EFF has filed an amicus brief in the case.

Our brief urges the district court to dismiss the case as soon as possible. Trademark is supposed to be about helping consumers identify the sources of the products they buy, not controlling criticism. That’s why our brief asks the court to use a test that’s more protective of speech than what’s applied in most trademark cases. This test, known as the Rogers test, has been adopted by many courts (but not yet this one) for cases where someone is using a trademark as part of an expressive work, rather than just as a brand name. We explain to the court that the Rogers test is an important First Amendment safeguard in part because it makes it easier to throw out meritless trademark claims before the most expensive parts of litigation, allowing more speakers to confidently stand up for their rights.

Our brief goes on to explain that First Amendment safeguards are especially important in cases like this one, where a plaintiff is seeking to control the use of a common term for its common meaning. Trademark law isn’t even supposed to extend to generic terms, and for good reason. Otherwise, we risk giving trademark owners power to control discussion and debate over entire topics.

It’s about time that a court shut down the LDS Church’s trademark bullying. We hope the court will do so here, while also taking the opportunity to endorse the Rogers test.

Artificial intelligence models are jacks of many trades, including writing, generating images, and creating 3D models. But they aren’t as helpful when it comes to testing robots or designs for vehicles in diverse environments, since they don’t understand physics as well as they do pixels or text.

To build an AI system that can reliably simulate a variety of physical scenarios, engineers need a range of physics data at a scale that isn’t yet feasible. That’s because it’s very time-consuming to get neural networks just a few data points they can understand. They rely on algorithms called “numerical solvers” to calculate physical properties at different points of a 3D shape. It’s a thorough process, but it takes so long that it limits how much data you’ll have to, say, test if your plane designs are safe and aerodynamic.

A new pre-training approach known as “GeoPT,” deveoped by researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and Tsinghua University, gives simulation models a chance to learn physics in a broader, more efficient way. It virtually reenacts everyday mechanical interactions in 3D, showing how particles stop when reaching some part of an object. These simulations give the models a sense of how physics works, helping them model the real world more accurately, reach peak performance twice as fast, and train on up to 60 percent less data compared to leading models.

Soon, the project could help engineers predict how vehicles (like cars and planes), everyday items (including chairs and containers), and robots respond to various physical elements, such as wind, water, and collisions. The researchers believe their work could also be a step toward a physics foundation model, a backbone system trained on lots of data that can help AI tools generalize to different tasks.

“We believe physics is the third modality for AI models, after text and pixels,” says MIT PhD student and CSAIL researcher Minghao Guo, a co-lead author on a paper introducing GeoPT. “Our general-purpose model has the versatility to help build a world model for physics. Many models, such as those that generate robotics data and videos, are already well-versed in textual and visual data, but with physical accuracy, they’ll get more-realistic results.”

Easy to use

To use GeoPT, users simply upload 3D models of objects like battleships, passenger airplanes, and trucks, and specify the direction and speed of the force they want to simulate. The result is a kind of heat map showing how the object will be affected in different places. If you know the speed and direction (velocity) of the force you’re looking to simulate, you can capture it in GeoPT. This comes in handy when you want to simulate things like how a car would look after crashing into a wall, the ways light bounces around objects, and whether a boat stays afloat over turbulent waves.

But how does GeoPT “get” physics so well? Its knowledge comes from “synthetic dynamics,” a series of interactions between small particles and complex 3D shapes. GeoPT studied 1.3 million samples of synthetic dynamics, in which tiny spheres moved at various speeds and angles until stopping at a certain point on the object.

These particles basically “stick” to an object once they make contact, instead of moving through or bouncing off. Picture learning about physical interactions using marbles and action figures — similarly, simulation models can use synthetic dynamics to gain a feel for physics before they train on labeled data.

Industry success

The researchers found that GeoPT was particularly skilled at simulating industrial scenarios, as it outperformed state-of-the-art simulation models across benchmarks. The common thread: It reached peak performance faster than other tools, while needing significantly fewer labeled data.

On a dataset of complex 3D shapes and their responses to wind currents and surface pressure, for example, GeoPT surpassed state-of-the-art models in speed, accuracy, and efficiency. It had similar triumphs in speed and accuracy in capturing how fighter jets responded to wind. When GeoPT tested how the hull of a boat handled both air and waves, it required 60 percent fewer labeled data to capture both physical forces and reached peak accuracy four times faster than top baselines.

The system even succeeded at simulating how different types of cars look after colliding with another object. It correctly predicted how 3D vehicles would deform while using less data than state-of-the-art baselines. Likewise, its simulations of how light would pass through what was essentially a toy rabbit were accurate, despite never training on that 3D model or light physics beforehand.

“If your model performs well on industrial benchmarks, that means it can solve the hardest physics tasks,” says co-lead author Haixu Wu, an MIT postdoc and CSAIL researcher. “GeoPT was making high-fidelity simulations with over 100 million mesh points in seconds. This could make the tool extremely helpful for engineers hoping to test out blueprints for vehicles without needing to run so many physical experiments.”

The researchers add that their system is only a preview of the kind of physics world model they’ve been working toward. The team hopes to scale up their system, training on even more shapes and simulating more complex physical phenomena. For example, a more in-depth approach could help model weather patterns, test out different materials, and generate realistic videos.

“Using synthetic dynamics data is an exciting paradigm for imbuing physics into foundation models,” says Fei Sha, AI research scientist at Meta, who wasn’t involved in the research. “It challenges the traditional wisdom that physics and geometry are necessarily entangled in computation, and one must acquire costly and specialized data. The demonstrated success in a wide range of application domains leads us to this important juncture: We are ready to build physics foundation models, now and fast."

Wu and Guo wrote the paper with MIT CSAIL colleagues including Zongyi Li, a postdoc in the lab; Zhiyang (Frank) Dou, a CSAIL affiliate and MIT PhD student in electrical engineering and computer science (EECS); Kaiming He, a principal investigator in the lab, associate professor of EECS, and a distinguished scientist at Google DeepMind; and senior author Wojciech Matusik, the Joan and Irwin M. (1957) Jacobs Professor of EECS and a CSAIL principal investigator. Tsinghua University Associate Professor Mingsheng Long was also a co-author. The team presented the paper at the International Conference on Machine Learning in July.

The researchers’ work was supported, in part, by Neural Modular Physics Twin for Robotics.

On Earth, the processes behind navigation are made nearly invisible by global positioning satellites (GPS). In cislunar space — the region between Earth and the moon — spacecraft do not have that kind of always-available positioning service. Missions beyond geosynchronous Earth orbit still rely heavily on NASA's Deep Space Network (DSN), an accurate but limited Earth-based international array of radio antennas shared across many missions and nations.

Because all DSN sites are located on Earth, their separation is small compared with the scale of cislunar space, which limits the angular baselines available for orbit determination. Therefore, precisely estimating orbits for distant spacecraft can take hours, and DSN supports only a few missions at a time. In addition, DSN requires user spacecraft to actively emit signals for measurement, unlike GPS, which passively sends data for users to receive.

The Laser Communications Group and Advanced Capabilities and Technologies Group at MIT Lincoln Laboratory are developing a concept called the Light High-Orbit Utility Signal Emitter (LightHOUSE) to help overcome these limitations. LightHOUSE would use a small constellation of satellites in high-altitude orbits as cooperative optical beacons. These beacons would exchange timing and communication signals with user spacecraft and use imaging against the stellar background to estimate each spacecraft's three-dimensional position and velocity. By providing timely, independent navigation data across cislunar space, LightHOUSE could reduce the need for corrective maneuvers, preserve spacecraft propellant, lessen the burden on onboard navigation sensors, and ease demand on existing ground-based systems.

“Satellites in cislunar space have limited access to support resources, even though orbits at and beyond the geosynchronous belt are increasingly important for various missions,” says Aaron Greenberg, a technical staff member in the Laser Communications Group. “The moon is reemerging as a strategic priority for national security. Nearly all space missions require some degree of precision navigation and timing, but no global positioning system exists in this domain. Here is where LightHOUSE is intended to step in, expanding critical and reliable communication and navigation services across this vast region.”

LightHOUSE would use free-space optical communications — laser links through space — rather than relying solely on radio-frequency systems. The concept builds on laboratory work demonstrated through NASA-sponsored programs such as TBIRD and O2O, as well as the Optical Time Transfer for Resilient Satellite Communications Networks project led by the Laser Communications Group with funding from the laboratory's internally administered R&D portfolio in optical systems technology.

“This concept hinges on a cooperative ranging capability enabled by free-space optical communications,” says Timothy Yarnall, an associate leader of the Laser Communications Group. “This technology area is one in which the laboratory is a global leader, as evidenced by the recent O2O success during Artemis II. The laboratory's experience with radiation hardening of digital focal plane array technology will also enable the sensitive receivers and star cameras — like the camera built by the Advanced Imager Technology Group for NASA's Psyche mission — that this concept relies upon.”

LightHOUSE beacons would be based in ultrahigh orbits, up to roughly 1 million miles in altitude. These high orbits replicate the angular diversity of GPS signals for users across cislunar volumes. They would also allow communication with spacecraft on the far side of the moon as viewed from Earth, preventing blackouts like the 40-minute period when Artemis II passed behind the moon.

Borrowing from the GPS philosophy, LightHOUSE is designed to place most of the technical burden on the beacon satellites, rather than on user spacecraft. The beacons would carry telescopes with tens-of-centimeter diameters and laser transmitters in the tens-of-watts range, while users would need only centimeter-scale apertures and tens-of-milliwatt lasers. The central engineering challenge is making that asymmetry work across cislunar space.

“From a design perspective, a major challenge will be making these services as easily accessible as possible to all potential users. The designed systems would be highly asymmetric, with LightHOUSE beacons taking on most technological and operational demands necessary to close links over the entire cislunar domain,” says Seth Trotz, a senior staff member in the Advanced Capabilities and Technologies Group.

Obtaining precise position measurements over such distances — combining optical communications with high-resolution imaging when beacons and user spacecraft are more than half a million miles from Earth — is itself a significant technical hurdle.

The team is now refining the system concept through analysis, simulation, and laboratory experimentation. In the near term, they plan to publish a detailed architecture for providing navigation data to LightHOUSE users. Longer term, the goal is to make navigation beyond geosynchronous altitudes routine, reliable, and accessible for a broad range of users, supporting Artemis and the growing wave of missions to follow in cislunar space.

This work is sponsored by the undersecretary of war for research and engineering through the laboratory's internally administered R&D portfolio in sensing and communications. A full-scale system would require substantial investment, potentially on the order of hundreds of millions of dollars; for comparison, the operating budget of GPS is $1.8 billion per year, and a single DSN dish costs roughly $85-100 million.

Access to accurate information about reproductive and maternal health can be critical. But on Meta's platforms, simply talking about prescription medication, abortion care, or one's own medical experiences can be enough to trigger content removals and account restrictions.

That's why EFF recently submitted a public comment to the Meta Oversight Board in its consideration of a case involving an Instagram post about prescription drugs during pregnancy and childbirth. The case touches upon a topic we’ve been documenting for some time; last year we collected stories from individuals who had experienced censorship of reproductive health information on various platforms. Meta in particular stood out: Its moderation systems routinely fail to distinguish between prohibited drug transactions and legitimate discussion of medications, including educational information and people's firsthand experiences with healthcare.

Through our Stop Censoring Abortion project, EFF collected nearly 100 submissions from healthcare providers, clinics, educators, advocates, researchers, and others whose reproductive health content had been removed or suppressed by social media platforms. What we found was alarming: systemic over-enforcement, confusing policies, arbitrary takedowns, sudden account bans, de-ranking, and appeals that too often went nowhere.

Talking About Medication Isn't the Same as Selling It

In almost every case we reviewed, the censored posts and accounts did not actually violate the platforms' stated rules. Meta frequently cited its Restricted Goods and Services policy, which prohibits attempts to buy, sell, trade, donate, gift, or request pharmaceutical drugs. But the content EFF documented overwhelmingly consisted of factual or educational information—not attempts to sell or distribute drugs.

The consequences were significant. For example, the Miscarriage+Abortion Hotline had its Instagram account restricted and posts removed even though it was providing information about legally obtaining medication rather than offering pharmaceuticals for sale. Red River Women's Clinic and the RISE reproductive health research center at Emory University had accounts locked after posting about mifepristone.

Other users reported having their content quietly de-ranked or “shadowbanned,” limiting its reach without giving them meaningful notice or recourse. We believe educational content and people's experiences involving reproductive healthcare and medication should not be suppressed in this way.

And when Meta gets these decisions wrong, the appeals process too often fails to fix them. In several cases EFF documented, accounts were restored only after journalists drew attention to the problem or someone with a personal connection inside Meta intervened. A moderation system shouldn't require knowing the right person to get an erroneous decision reversed.

Meta Can—and Must—Do Better

Our submission calls on Meta to make five changes—the same five changes we asked for last year.

First, Meta should publish clear, understandable policies so users can know what content is permitted and what might result in removal, downranking, or account suspension. Second, those rules must be enforced consistently and fairly. Third, Meta must provide meaningful explanations for enforcement decisions, including what rule was violated and how users can appeal. Fourth, users need a functional appeals system that doesn't depend on insider access.

Finally, Meta should expand human review. Reproductive healthcare is precisely the sort of nuanced and context-dependent subject that automated moderation systems struggle to understand. As our research shows, automated systems can mistake education for drug sales, misinterpret terminology, overlook cultural and political context, and even classify legitimate advocacy as dangerous content. Human moderators should therefore play a greater role when automated systems flag sensitive healthcare information or political expression.

Meta has chosen to allow discussion of reproductive healthcare, including abortion, on its platforms. That commitment means little if its moderation systems nevertheless prevent people from accessing or sharing that information.

At a moment when reproductive rights are under attack around the world, the stakes are particularly high. Restricting access to essential healthcare information can have profound consequences, especially for people who already face barriers to reproductive care.

Users deserve a system in which rules aren't applied arbitrarily, appeals actually work, and vital health information isn't silenced because an automated system failed to understand its context. Meta can—and must—do better.

You can read our comment in full below.

This is good:

Post-quantum cryptography is now one pip-install away for the entire Python ecosystem. With funding from the Sovereign Tech Agency, we implemented support for ML-KEM, the NIST-standard key-establishment primitive, and ML-DSA, the NIST-standard digital-signature primitive, in pyca/cryptography.

Remember, the reason to do this now is because there’s no emergency. And because you will make your systems crypto agile, which is always a good idea.

Some environmentalists are looking to rebuild a cordial relationship with Gov. Kathy Hochul as they look for alignment after a bitter fight over the state’s climate law.

Cameron Hamilton's arrival Monday as FEMA's first permanent administrator in Donald Trump's second term portends changes to disaster aid.

The California governor made his first public remarks detailing his priorities for this year on the issue.

The agency said it reached an agreement with groups affected by billions of dollars in frozen funds. Money will not flow while the case goes through the appeals process.

Gov. Spencer Cox has drawn support across the political spectrum for "Operation Gigawatt” as his state tries to plan for surging power demand.

Dry seasons have become a fixture of European farming. The EU’s crisis tools still treat them as a surprise.

The Maldives has championed itself as a country fighting its vulnerability to climate change, but the government has said sacrifices must be made at times in favor of development.

The Bureau of Reclamation said that water from Glen Canyon Dam would not be released “due to significant hydropower generation concerns under already strained system conditions."

Heavy rains have created a risk of severe flooding in Zhejiang province, and authorities have warned of severe landslide risks.

Nature Climate Change, Published online: 10 August 2026; doi:10.1038/s41558-026-02735-5

Author Correction: Enduring impacts of El Niño on life expectancy in past and future climates

Nature Climate Change, Published online: 10 August 2026; doi:10.1038/s41558-026-02724-8

The authors quantify the global emergence of hourly heat extremes (HHEs), highlighting exposure overlooked by daily metrics. HHEs will increase fourfold by the end of the century under high emissions, with low- and middle-income countries, as well as successive generations, facing disproportionate exposure.

In the last decade, scientists have shown that fusion energy can work, as a physical process. Next question: Can it work economically? 

A study co-authored by MIT professors Dennis Whyte and Andrew W. Lo proposes a framework for understanding what’s needed to make fusion energy commercially viable in the marketplace. The method considers the physical inputs needed to sustain controlled fusion energy production, as well as the cost of building power plants that can compete in energy markets.

“It’s all the things that come along with finding, allocating, and spending money at this scale,” says Whyte, a professor of nuclear science and engineering at MIT and a key driver of the field’s progress, who co-authored the paper. “This is critical to what we do. We should look at the economics. If we want this technology to actually be meaningful in the world economy, we have to start getting straight with ourselves about these topics.” 

The goal of the paper, Whyte says, is to create “this framework, where are all the economics are clear, and then we understand what it would mean” for any fusion energy power plant. 

Fusion energy harnesses the reaction that powers the stars: the fusion of light nuclei. It is often referred to as “plasma fusion,” as the fusion reactions generate fuel in a plasma state, often confined by magnets or initiated by powerful lasers. Whyte says the goal is to generate abundant energy while also offering society attractive safety, licensing, and siting options.

In 2022, researchers at the National Ignition Facility in Livermore, California, one of the U.S. national labs, achieved a reaction with positive energy gain. Venture funding has also poured into the field in recent years, although there are still many challenges regarding the construction of viable commercial fusion energy.

“It’s challenging to reduce complex scientific and engineering requirements to economic consequences,” Lo says. “But if we don’t do that, we’re not going to get the funding we need to achieve the impact we want.”

The open-access publication, “Criteria for the economic viability of fusion power plants,” appears online in the Journal of Fusion Energy. The authors are Whyte, who is the Hitachi America Professor of Engineering and a professor of nuclear science and engineering at MIT; Lo, who is the Charles E. and Susan T. Harris Professor and a professor of finance at the MIT Sloan School of Management; Rachel Bielajew, an analyst with Rutherford Energy Ventures and a researcher at MIT’s Plasma Science and Fusion Center; Maria Hancock and Riley Moeykens of Rutherford Energy Ventures; and Guinevere Shaw of Rutherford Energy Ventures and MIT’s Plasma Science and Fusion Center.

Whyte is a former head of MIT’s Department of Nuclear Science and Engineering and a former director of MIT’s Plasma Science and Fusion Center. He co-founded Commonwealth Fusion Systems, an MIT spinoff firm that is one of the leaders in the fusion industry. Whyte and Lo also co-founded Rutherford Energy Ventures, a consultancy and investment advisory firm, which is working with the U.S. Department of Energy’s Oak Ridge National Laboratory to build a consortium for new fusion research. 

10 parameters, any power plant

The framework Whyte and Lo propose in the paper has 10 parameters for evaluating the economic viability of a fusion energy power plant. Some of these are scientific and physical, dealing with the energy consumed and produced in a given plant. Most of the parameters are in the realm of engineering and economics, such as the costs of plant construction.

A key inspiration for the framework is the so-called Lawson Criterion, derived in the 1950s, which describes the combinations of temperature, plasma density, and energy confinement time that can produce net energy from the plasma due to fusion, regardless of its absolute power or volume. Specifically it calculates a “plasma Q,” which is the ratio of fusion power produced to the external power required to sustain the plasma.

“The Lawson Criterion describes the scientific success of energy gain from fusion plasmas, while our framework generally describes economic Q, which is the ratio of capital gained to that expended,” Whyte explains.

The parameters in the framework describe engineering features of the fusion power plant such as power density, the efficiency of converting fusion power into an economic product, and the durability of components used in the energy conversion, in addition to costing and market parameters that assess the expenses and returns from invested capital. Or, as Whyte puts it, the framework is centered on what it takes to achieve a net-positive economic return, “but applied to practical power plant design.” In parallel to plasma Q, the economic Q described in the framework must be greater than 1 for basic viability.

Researchers have tried a variety of methods for generating and containing fusion energy. The paper’s framework, Whyte emphasizes, is “completely agnostic to whatever fusion concept you use, because the physical reality of fusion is that you expend money to build the capability to produce fusion power.” And the parameters do not depend on the size of any reactor being built; the framework is set up so that any inputs can be scaled to a given project or power output. 

“It doesn’t matter whether the fusion power plant is a small or large, the bottom line is: In both cases you better have money coming out that exceeds the money going in, otherwise it’s not going to be around for very long,” Lo says. 

One source of motivation for the paper, Whyte and Lo say, is to underscore the importance of accounting for all costs in fusion research as rigorously as possible. While researchers will be highly aware of the costs of basic experiments, estimating the costs of a fusion reactor is a somewhat different matter, but something leaders in the field have to be increasingly oriented around.

Fixing a missing link

That is certainly the case, the authors note, as new rounds of funding enter the fusion energy industry. Just last week, Commonwealth Fusion Systems obtained a new billion-dollar round of funding support from investors; it hopes to open its first working power plant in the 2030s, in the state of Virginia. 

Lo acknowledges that there will be uncertainties and challenging decisions involved in the development of the very first commercial fusion reactor. If successful, though, the industry might follow the path of learning by doing that has been common in energy and other industries, helping plants become more economical over time. 

“This pattern of learning by doing exists in all deep technology sectors,” says Lo, noting that sequencing a human genome is a million times cheaper right now than it was about 25 years ago. “We’re going to see the same thing, but maybe not to the same degree, in fusion energy.”

Lo has long worked to develop ways for scientific research to gain financial support in biotechnology — and is launching a new MIT Sloan educational program, called CATAPULT, to provide more tools for people in any field of study to translate their research advances into products. 

When it comes to fusion, Lo says, “It’s pretty clear that economic viability is something we can start assessing now.” And while there might be thousands of particular decisions involved in building a commercial fusion plant, the authors think they have an overall approach that will let people quantify all that work. 

“When you’ve got a framework to evaluate it in a quantitative way, it tells you about the literal worth of making a particular design decision. That seems to me at this moment of fusion development absolutely critical, and what we’ve been missing,” Whyte says.

Nice video of the Arctic bobtail squid.

As usual, you can also use this squid post to talk about the security stories in the news that I haven’t covered.

Blog moderation policy.

Through data brokers, ICE is buying the information you provided to open a credit card.

A fight over USDA immigration and diversity policies held up $15 million in wildfire-prevention grants for an area in Washington state that’s now burning.