6 February 2026

Today’s Protostar is Mayank Kejriwal, whose work using AI to study human trafficking made him the runner-up for the inaugural ASU-Science Prize for Transformational Impact. But first, catch up on the latest science news, including a bonobo with an imagination and how a kidney disease drug could help people conceive.

Computer Science | News from Science

Quantum encryption scheme neutralizes backdoor threats, paving way for secure internet

Powerful quantum computers may soon be able to crack the secret codes that safeguard sensitive information on the internet. Luckily, physicists have a way to fight quantum with quantum: a theoretically foolproof encryption protocol that relies on the magic of entangled quantum particles. The scheme closes loopholes in existing methods that leave hardware vulnerable to hacking. A team in China has now demonstrated the technique on metropolitan scales, an advance that could pave the way for a secure quantum internet in which trust is taken for granted.

“After so many years of effort … it’s no longer a theoretical idea,” said Jian-Wei Pan, who led the study.

Entanglement is fundamentally monogamous: If two particles are maximally entangled then no other particle can join the connection. If the sender and receiver entangle a pair of particles across the network, they can each perform tests that confirm the particles’ properties are strongly correlated, well beyond chance. After this “handshake,” they can be sure they’re the only ones on the channel. Then, other measurements on the entangled particles can establish a key that nobody can decipher.

At each end of a coiled fiber optic line, Pan and his colleagues trapped a rubidium atom and stabilized it using lasers. They then coaxed the atoms to emit single photons, which entangled the separated atoms. After collecting data for 26 days, they showed they had the statistics to establish and share a key across 11 kilometers, they report this week in Science.

For now, the scheme is expensive, complicated, and impractical. But physicist Tracy Northup says the advance is a critical step toward a secure global quantum internet. “I see it as taking advantage of these fundamental aspects of quantum physics in ways that can make our lives better.”

Read the Science paper

Read the full story

Comparative Cognition | News From Science

Imagination isn’t just for humans, this famous ape shows

Kanzi’s imagination wowed researchers. Ape Initiative

To primatologists around the world, Kanzi was a star. After years of training with sign language and a specialized keyboard, the bonobo had gained an unusually sophisticated understanding of human words, making him the focus of decades of research into nonhuman primates’ language abilities. But in March 2025, researchers wanted to test something different: Kanzi’s imagination.

After setting up versions of a child’s tea party with the bonobo, the team found that Kanzi could reliably keep track of glasses “filled” with imaginary juice or grapes and distinguish between glasses filled with real juice and make-believe juice. The discovery, reported yesterday in Science , shows that imagination is not uniquely human and suggests it may have arisen millions of years ago, possibly among the common ancestors of humans, bonobos, and chimpanzees.

Both captive and wild apes have dabbled in alternate realities. Anecdotally, scientists have reported apes occasionally playing pretend, cradling logs in their arms as if they were children or mimicking the motions of playing with toys. Research going back decades also shows apes can plan ahead and consider the beliefs of other individuals. But until Kanzi, there had never been reproducible evidence of an ape’s ability to engage with make-believe objects.

The finding also represents the final piece of Kanzi’s scientific legacy. Soon after these experiments, the bonobo died at the age of 44. “Kanzi left us that one more little piece of the puzzle before he went,” said the study’s lead author Amalia Bastos.

read the Science Paper

Read the full story

Medicine | Science

Could a kidney disease drug also treat infertility?

Primary ovarian insufficiency (POI), in which the ovaries don’t mature and release eggs properly, affects about 3% of women below age 40; it typically causes infertility. But a drug already approved to treat kidney disease may one day help some of these people get pregnant, a research team reports this week in Science.

In studies of mouse ovarian tissue and live mice, the drug finerenone kick-started the growth of tiny egg-containing sacs, or follicles—a process that’s often defunct in people with POI. And in a small clinical trial, people with POI who were given the drug went on to produce mature eggs that could be harvested. Some of the eggs were then successfully fertilized in the lab; none of the resulting embryos have yet been implanted.

“Although these clinical results are promising, they are preliminary ,” noted Francesca Duncan and Elnur Babayev in a related Perspective. The trial didn’t have a control group, so it’s difficult to disentangle how much finerenone helped—on rare occasions people with POI can become pregnant without any intervention. And more studies are needed to better understand finerenone’s effect on ovaries and to rule out any harmful effects on pregnant people or fetuses. Still, the approach appears promising, said Suzannah Williams, an ovarian physiologist who was not involved in the study. “ The benefit of this is they’ve got a drug that is already past FDA, which I think is a huge bonus.”

Read the Science paper and related PERSPECTIVE

Read the full story

SPONSORED

The latest advances in cancer immunotherapy will be presented at AACR IO

This must-attend event will feature cutting-edge research across the spectrum of basic, translational, and clinical science in immunology, inflammation, and cancer immunotherapies, presented by leading experts from both academia and industry. 

Learn More

Protostar

PHOTO USC Information Sciences Institute

Mayank Kejriwal

Research Associate Professor, University of Southern California

Kejriwal, M. Turning the web against sex traffickers: Scalable artificial intelligence is harnessed for investigative search. Science 391 (2026). 10.1126/science.aee1347

Mayank Kejriwal’s career in research was inspired by Star Trek. “I never planned on becoming [a scientist],” he explains to ScienceAdviser. “What I remember is watching reruns of Star Trek Voyager … They were really early in the morning, so I would wake up early and watch it before going to school. I remember being influenced by the computer—I realize now the computer on Star Trek, you could talk to it and it does things; it’s just like a large language model. At the time it seemed magical, but now we have technology like that.”

His fascination with computers grew as he did, and he decided to take a class on artificial intelligence as an undergraduate student. “The class was really amazing,” he recalls. “It really opened my mind up.” He went on to do a research project over the summer, and he was hooked on computer science and AI. He initially planned to go into industry instead of academia, but he was invited by researchers at USC to give a talk and learned about a project they were working on to leverage AI to fight human trafficking. “I felt like I had to do this,” he says.

And he did—research that made Kejriwal the runner-up for the inaugural ASU-Science Prize for Transformational Impact. ScienceAdviser chatted with him about the work and the honor; that interview is below, edited for clarity and brevity.

How did your work on this issue begin?
When I was invited to give a talk for an open position by influential scientists at the USC Information Sciences Institute, they told me about this project, which was being funded by DARPA at the time. So I joined the institute.

I didn’t expect it to have the kind of influence it had on me. I ended up thinking a lot like a social scientist and learning a lot about human trafficking and just how prevalent the problem is, even in developed countries—and how technology has made the problem worse, but also, how we can use technology to make the problem better. That was the premise of the program, which ended after three or four years, as all DARPA programs do. But I just kept going, getting donors involved and working with stakeholders. And one way or the other, the project has continued over the last 10 years.

I think that is one of the reasons why I was chosen as runner-up for this award—because the committee recognized not just the problem itself, but that we’ve kept going, even though there were times when we didn’t have money for it, or there was no momentum. We always found a way to keep it going. Now we are at the stage where a lot of exciting things will start happening, because some of our stakeholders have trusted us with data, with insights, and we want to be very good stewards of what they have given us.

Describe the work you’ve done so far, and what’s on the horizon.
So there was this website which was called Backpage.com. It was kind of like Craigslist, where people advertise things like used furniture and all of that, and it was a very big website back almost 10 years ago. But unfortunately, about 80% or 90% of Backpage’s profits came from sex advertisements, and a significant fraction of that was human trafficking, and we now know that they turned a blind eye to all of that. In Between 2018 and 2019, the Trump administration passed a law, and under that law, they were able to prosecute Backpage. The Backpage Chief Executive Officer and other people who are very highly ranked decided to cooperate with the authorities. And the federal agencies then came to me a few years later, and they were like, “We have all of this data from Backpage. Can you help us to find victims of child sex trafficking?”

So we help them, and then we had that data to conduct scientific studies on. Now, we have to be very careful. We can only do it in aggregate. We have to make sure that no private data is ever identified. There are certain questions which are ethically problematic, so we would not do those, even though they are important. And so, what we are trying to do now is to conduct an international study of the sex work market. We call it the sex work market, because not all sex workers are trafficked, and we are very careful to recognize that. But there is no question that that is where it begins. And there are certain regions of the world where this is a very big problem. We are trying to understand: What can we learn about it? What are the numbers? How many people and what ethnicities? What does the pricing structure look like? What does the network structure look like? These are all questions that we are now trying to answer scientifically.

We already have a couple of papers that we have written on the Southeast Asian region as well as the Caribbean region. They are under peer review right now, but there will be more like that coming out both this year and the next year. Our goal is that by the time we are done, which we assume would be like end of next year, there will be a whole range of papers that will share insights on the sex work market that have not, to our knowledge, been uncovered ever before. And we are able to do that because we got this data from law enforcement and federal agencies who trusted us to do the right thing and study the data properly.

Is there anything you’d like to add?
I feel really lucky and honored to be given this award. I think that this award is unusual, because they were specifically looking at how you influenced policy or the broader society with your work, how you had a transformative impact, not just technical merit. Of course, we want that as scientists—to have that invention or that discovery that will change everything. But sometimes it is easy to ignore the inspection of problem domains like this. There’s no invention that will cure human trafficking, unfortunately. Even if we get rid of the internet, it will still go on. It’s been going on for thousands of years. So, there’s no invention, there’s no theorem that will solve this problem. In these kinds of wicked problems in social science—we call these wicked problems, because they are problems where every time you propose a solution to the problem, you might be opening up a Pandora’s box and creating more problems—we have to be very careful. It’s very complex. It’s very nonlinear. We have to take all our stakeholders into account.

I hope that awards like these and work like this really communicates that to scientists—that we really should engage with stakeholders, and figure out what they want, what they need. It’s not just thinking about problems in a very clean and abstract way, but getting into the messiness of it. It is important to get into the messiness to have real impact. I feel like it’s looked at as a bug, but messiness is a feature. It’s what makes the problem interesting and worth working on.

Even if we had no papers or awards, to know that the work has helped actual human trafficking victims, and just talking to people who have been touched by the work—it’s an indescribable feeling. I did not know that I would experience that as a scientist, but I would not exchange that for anything in the world. It’s a privilege to work on this problem, and I hope that I can keep working on it all my life. I mean, unfortunately, the problem is not going away—I wish that it would go away tomorrow so that I could stop working on it. But it’s not going away anytime soon, and so all we can really do is try our best to make it better.

Read the Essay

podcast

Engineering safer football helmets, and the science behind drug overdoses

By SARAH CRESPI, Adrian Cho | 5 February 2025

Et Cetera

More allegations against BHP

The Bandim Health Project (BHP) has recently become influential in U.S. health policy—but in a new paper, researchers point out key data are missing for 10 trials conducted in West Africa, which together involved tens of thousands of children. “It is a red flag when people don’t report the primary result of a randomized trial,” said one expert. “We have an ethical obligation to participants in randomized trials to report the results for the research questions that those trials were designed to answer.”

Vaccine Paper | Read more at ScienceInsider

Hijack and hide

Lung cancer tumors hack into nearby neurons and send signals to quell the immune system, allowing the invasive cancers to grow unchecked. When researchers knocked out these neurons in mice, the tumors grew up to 50% less. “The tumor hijacks the signaling axis and uses it for its own purpose,” one cancer immunologist explained.

Nature Paper | Read more at Nature

Padded diagnostics

The menstrual blood collected on a sanitary pad contains valuable intel. Researchers in China were able to detect HPV, the virus that can cause cervical cancer, in samples from pads just as well as in clinically collected ones. “I believe that the more clinically effective and validated tools we have to reach a broader proportion of people who persistently remain underscreened, the better,” one expert said.

The BMJ Paper | Read more at Scientific American

Last but not least

I have a newfound appreciation for football helmets after reading News from Science Staff Writer Adrian Cho’s incredible feature on their physics. And just in time to watch a bunch of them crash together this weekend... This New England girl simply has to root for her team. Go Patriots!

Christie Wilcox, Editor, ScienceAdviser

With contributions from Zack Savitsky, Michael Greshko, and Kelly Servick

Do you have a burning science question you can’t seem to find a good answer for? Submit it to Ask Science! Selected questions will receive responses from Science editors right here in ScienceAdviser.

Have feedback on this newsletter? Let us know what you think using this form or drop us a note at ScienceAdviser@aaas.org.

Want more? Catch up on past issues of ScienceAdviser.

If you were forwarded this newsletter by a friend, you can subscribe for free here.

To ensure ScienceAdviser lands in your inbox, consider taking a moment to add scienceadviser@aaas.sciencepubs.org as a trusted sender or contact in your email client. These instructions provide more information on whitelisting ScienceAdviser based on email client.