A recent policy forum paper published in Science describes how large groups of AI-generated personas can convincingly imitate human behavior online. These systems can enter digital communities, participate in discussions, and influence viewpoints at extraordinary speed. Unlike earlier bot networks, these AI agents can coordinate instantly, respond to feedback, and maintain consistent narratives across thousands of accounts.

How AI personas mimic real people online

Rapid progress in large language models and multi-agent systems has made it possible for a single operator to manage vast networks of AI “voices.” These personas can appear authentic, adopt local language and tone, and interact in ways that feel natural to other users.

They are also capable of running millions of small-scale experiments to determine which messages are most persuasive. This allows them to refine their communication strategies in real time and generate what appears to be widespread public agreement. In reality, that consensus is artificially created and designed to influence political discussions.

Deepfakes and fake news signal early risks

Although fully developed AI swarms are still largely theoretical, researchers say there are already warning signs. These include AI-generated deepfakes and fake news outlets that have influenced recent election conversations in countries such as the United States, Taiwan, Indonesia, and India, according to UBC computer scientist Dr. Kevin Leyton-Brown.

At the same time, monitoring organizations have identified pro-Kremlin networks spreading large volumes of online content. This activity is believed to be aimed at shaping the data used to train future AI systems, potentially influencing how those systems behave and what information they prioritize.

Experts warn of growing impact on democracy

Looking ahead, experts believe AI swarms could significantly affect the balance of power in democratic societies. Dr. Leyton-Brown cautioned that these systems are likely to change how people trust information online. “We shouldn’t imagine that society will remain unchanged as these systems emerge. A likely result is decreased trust of unknown voices on social media, which could empower celebrities and make it harder for grassroots messages to break through.”

Researchers suggest that upcoming elections may serve as a critical test for this technology. The key challenge will be recognizing and responding to these AI-driven influence campaigns before they become too widespread to control.

The primary drivers behind this trend include intensive groundwater extraction, a decline in sediment carried by rivers, and rapid urban development.

Global Mapping Reveals Widespread Delta Sinking

This research offers the first detailed, high-resolution analysis of elevation loss across 40 river deltas around the world. The project was led by Leonard Ohenhen, a former Virginia Tech graduate student who is now an assistant professor at the University of California, Irvine. The work was overseen by Virginia Tech geoscientists Manoochehr Shirzaei and Susanna Werth.

Results show that almost every delta studied contains areas where the land is dropping faster than nearby sea levels are rising. In 18 of the 40 deltas, this downward movement, known as subsidence, already exceeds local sea-level rise. That trend is increasing near-term flood risk for more than 236 million people.

Satellite Data Tracks Elevation Loss Across Continents

Researchers used advanced satellite radar systems to measure changes in surface elevation across deltas on five continents. The resulting high-resolution maps capture changes at a scale of 75 square meters per pixel, allowing scientists to detect localized patterns of sinking.

Several major deltas are experiencing especially rapid elevation loss, including those of the Mekong, Nile, Chao Phraya, Ganges-Brahmaputra, Mississippi, and Yellow rivers.

“In many places, groundwater extraction, sediment starvation, and rapid urbanization are causing land to sink much faster than previously recognized,” Ohenhen said.

In some areas, the rate of sinking is more than double the current global pace of sea-level rise.

Human Activity Driving Accelerated Subsidence

“Our results show that subsidence isn’t a distant future problem — it is happening now, at scales that exceed climate-driven sea-level rise in many deltas,” said Shirzaei, co-author and director of Virginia Tech’s Earth Observation and Innovation Lab.

The study identifies groundwater depletion as the strongest overall factor linked to delta subsidence, although the main cause varies by region.

“When groundwater is over-pumped or sediments fail to reach the coast, the land surface drops,” said Werth, who co-led the groundwater analysis. “These processes are directly linked to human decisions, which means the solutions also lie within our control.”

The research was supported by the National Science Foundation, the Department of Defense, and NASA.

Their work solves a long-standing geological puzzle known as the “Dolomite Problem.” Dolomite is a widespread mineral found in iconic locations such as the Dolomite mountains in Italy, Niagara Falls and Utah’s Hoodoos. It is abundant in rocks older than 100 million years, yet it is rarely seen forming in more recent environments.

“If we understand how dolomite grows in nature, we might learn new strategies to promote the crystal growth of modern technological materials,” said Wenhao Sun, the Dow Early Career Professor of Materials Science and Engineering at U-M and the corresponding author of the paper published in Science.

Why Dolomite Growth Is So Slow

The key breakthrough came from understanding what disrupts dolomite as it forms. In water, minerals typically grow as atoms attach in an orderly way to the surface of a crystal. Dolomite behaves differently because its structure is made of alternating layers of calcium and magnesium.

As the crystal grows, these two elements often attach randomly instead of lining up correctly. This creates structural defects that block further growth. The result is an extremely slow process. At that rate, forming a single well-ordered layer of dolomite could take up to 10 million years.

Nature’s Built-In Reset Mechanism

The researchers realized that these defects are not permanent. Atoms that are out of place are less stable and more likely to dissolve when exposed to water. In natural environments, cycles such as rainfall or tidal changes repeatedly wash away these flawed areas.

Over time, this process clears the surface so new, properly arranged layers can form. Instead of taking millions of years for a single layer, dolomite can gradually build up in far shorter intervals. Over long geological periods, this leads to the large deposits seen in ancient rock formations.

Simulating Crystal Growth at the Atomic Level

To test their idea, the team needed to model how atoms interact as dolomite forms. This requires calculating the energy involved in countless interactions between electrons and atoms, which is usually extremely demanding in terms of computing power.

Researchers at U-M’s Predictive Structure Materials Science (PRISMS) Center developed software that simplifies this challenge. It calculates the energy for certain atomic arrangements and then predicts others based on the symmetry of the crystal structure.

“Our software calculates the energy for some atomic arrangements, then extrapolates to predict the energies for other arrangements based on the symmetry of the crystal structure,” said Brian Puchala, one of the software’s lead developers and an associate research scientist in U-M’s Department of Materials Science and Engineering.

This approach made it possible to simulate dolomite growth over timescales that reflect real geological processes.

“Each atomic step would normally take over 5,000 CPU hours on a supercomputer. Now, we can do the same calculation in 2 milliseconds on a desktop,” said Joonsoo Kim, a doctoral student of materials science and engineering and the study’s first author.

Lab Experiment Confirms the Theory

Natural settings where dolomite still forms today often experience cycles of flooding followed by drying, which supports the team’s theory. However, direct experimental evidence was still needed.

That evidence came from Yuki Kimura, a professor of materials science at Hokkaido University, and Tomoya Yamazaki, a postdoctoral researcher in his lab. They used an unusual property of transmission electron microscopes to recreate the process.

“Electron microscopes usually use electron beams just to image samples,” Kimura said. “However, the beam can also split water, which makes acid that can cause crystals to dissolve. Usually this is bad for imaging, but in this case, dissolution is exactly what we wanted.”

The team placed a small dolomite crystal in a solution containing calcium and magnesium. They then pulsed the electron beam 4,000 times over two hours, repeatedly dissolving the defects as they formed.

After this process, the crystal grew to about 100 nanometers, or roughly 250,000 times smaller than an inch. That growth represented around 300 layers of dolomite. Previous experiments had never produced more than five layers.

Implications for Modern Technology

Solving the Dolomite Problem does more than explain a geological mystery. It also offers insight into how to control crystal growth in advanced materials used in modern technology.

“In the past, crystal growers who wanted to make materials without defects would try to grow them really slowly,” Sun said. “Our theory shows that you can grow defect-free materials quickly, if you periodically dissolve the defects away during growth.”

This concept could help improve the production of semiconductors, solar panels, batteries and other high-performance technologies.

The research was funded by the American Chemical Society PRF New Doctoral Investigator grant, the U.S. Department of Energy and the Japanese Society for the Promotion of Science.