Science & Technology

5 ways to make AI more trustworthy

Amber Elise Carlson — Tue, 21 Oct 2025 14:52:17 +0000

5 ways to make AI more trustworthy Amber Elise Carlson Tue, 10/21/2025 - 08:52

Amber Carlson

A Waymo self-driving taxi. (Credit: Adobe Stock)

Self-driving taxis are sweeping the country and will likely start service in Colorado in the coming months. How many of us will be lining up to take a ride?

That depends on our level of trust, says Amir Behzadan, a professor in the Department of Civil, Environmental and Architectural Engineering, and a fellow in the Institute of Behavioral Science (IBS) at ̽��Ƶ.

He and his team of researchers in the Connected Informatics and Built Environment Research (CIBER) Lab at ̽��Ƶ are unearthing new insights into how the artificial intelligence (AI) technology we might encounter in daily life can earn our confidence. They’ve created a framework for developing trustworthy AI tools that benefit people and society.

And in a new paper in the journal AI and Ethics, Behzadan and his Ph.D. student Armita Dabiri drew on that framework to create a conceptual AI tool that incorporates the elements of trustworthiness.

“As a human, when you make yourself vulnerable to potential harm, assuming others have positive intentions, you’re trusting them,” said Behzadan. “And now you can bring that concept from human-human relationships to human-technology relationships.”

How trust forms

Behzadan studies the building blocks of human trust in AI systems that are used in the built environment, from self-driving cars and smart home security systems to mobile public transportation apps and systems that help people collaborate on group projects. He says trust has a critical impact on whether people will adopt and rely on them or not.

Trust is deeply embedded in human civilization, according to Behzadan. Since ancient times, trust has helped people cooperate, share knowledge and resources, form communal bonds and divvy up labor. Early humans began forming communities and trusting those within their inner circles.

Mistrust arose as a survival instinct, making people more cautious when interacting with people outside of their group. Over time, cross-group trade encouraged different groups to interact and become interdependent, but it didn’t eliminate mistrust.

We can see echoes of this trust-mistrust dynamic in modern attitudes toward AI, says Behzadan, especially if it’s developed by corporations, governments or others we might consider “outsiders.” So what does trustworthy AI look like? Here are five main takeaways from Behzadan’s framework.

1. It knows its users.

Amir Behzadan

Many factors affect whether—and how much—we trust new AI technology. Each of us has our own individual inclination toward trust, which is influenced by our ences, value system, cultural beliefs, and even the way our brains are wired.

“Our understanding of trust is really different from one person to the next,” said Behzadan. “Even if you have a very trustworthy system or person, our reaction to that system or person can be very different. You may trust them, and I may not.”

He said it’s important for developers to consider who the users are of an AI tool. What social or cultural norms do they follow? What might their preferences be? How technologically literate are they?

For instance, Amazon Alexa, Google Assistant and other voice assistants offer simpler language, larger text displays on devices and a longer response time for older adults and people who aren’t as technologically savvy, Behzadan said.

2. It’s reliable, ethical and transparent.

Technical trustworthiness generally refers to how well an AI tool works, how safe and secure it is, and how easy it is for users to understand how it works and how their data is used.

An optimally trustworthy tool must do its job accurately and consistently, Behzadan said. If it does fail, it should not harm people, property or the environment. It must also provide security against unauthorized access, protect users’ privacy and be able to adapt and keep working amid unexpected changes. It should also be free from harmful bias and should not discriminate between different users.

Transparency is also key. Behzadan says some AI technologies, such as sophisticated tools used for credit scoring or loan approval, operate like a “black box” that doesn’t allow us to see how our data is used or where it goes once it’s in the system. If the system can share how it’s using data and users can see how it makes decisions, he said, more people might be willing to share their data.

In many settings, like medical diagnosis, the most trustworthy AI tools should complement human expertise and be transparent about their reasoning with expert clinicians, according to Behzadan.

AI developers should not only try to develop trustworthy, ethical tools, but also find ways to measure and improve their tools’ trustworthiness once it is launched for the intended users.

3. It takes context into account.

There are countless uses for AI tools, but a particular tool should be sensitive to the context of the problem it’s trying to solve.

In the newest study, Behzadan and co-researcher Dabiri created a hypothetical scenario where a project team of engineers, urban planners, historic preservationists and government officials had been tasked with repairing and maintaining a historical building in downtown Denver. Such work can be complex and involve competing priorities, like cost effectiveness, energy savings, historical integrity and safety.

The researchers proposed a conceptual AI assistive tool called PreservAI that could be designed to balance competing interests, incorporate stakeholder input, analyze different outcomes and trade-offs, and collaborate helpfully with humans rather than replacing their expertise.

Ideally, AI tools should incorporate as much contextual information as possible so they can work reliably.

4. It’s easy to use and asks users how it’s doing.

The AI tool should not only do its job efficiently, but also provide a good user experience, keeping errors to a minimum, engaging users and building in ways to address potential frustrations, Behzadan said.

Another key ingredient for building trust? Actually allowing people to use AI systems and challenge AI outcomes.

“Even if you have the most trustworthy system, if you don't let people interact with it, they are not going to trust it. If very few people have really tested it, you can't expect an entire society to trust it and use it,” he said.

Finally, stakeholders should be able to provide feedback on how well the tool is working. That feedback can be helpful in improving the tool and making it more trustworthy for future users.

5. When trust is lost, it adapts to rebuild it.

Our trust in new technology can change over time. One person might generally trust new technology and be excited to ride in a self-driving taxi, but if they read news stories about the taxis getting in crashes, they might start to lose trust.

That trust can later be rebuilt, said Behzadan, although users can remain skeptical about the tool.

For instance, he said, the “Tay” chatbot by Microsoft failed within hours of its launch in 2016 because it picked up harmful language from social media and began to post offensive tweets. The incident caused public outrage. But later that same year, Microsoft released a new chatbot, “Zo,” with stronger content filtering and other guardrails. Although some users criticized Zo as a “censored” chatbot, its improved design helped more people trust it.

There’s no way to completely eliminate the risk that comes with trusting AI, Behzadan said. AI systems rely on people being willing to share data—the less data the system has, the less reliable it is. But there’s always a risk of data being misused or AI not working the way it’s supposed to.

When we’re willing to use AI systems and share our data with them, though, the systems become better at their jobs and more trustworthy. And while no system is perfect, Behzadan feels the benefits outweigh the downsides.

"When people trust AI systems enough to share their data and engage with them meaningfully, those systems can improve significantly, becoming more accurate, fair, and useful,” he said.

“Trust is not just a benefit to the technology; it is a pathway for people to gain more personalized and effective support from AI in return.”

Beyond the Story

Our research impact by the numbers:

$742 million in research funding earned in 2023–24
No. 5 U.S. university for startup creation
$1.4 billion impact of ̽��Ƶ's research activities on the Colorado economy in 2023–24

Follow ̽��Ƶ on LinkedIn

̽��Ƶ engineers have designed a framework to help technology developers create artificial intelligence people will actually want to use.

Traditional

White

Open-source software allows for efficient 3D printing with multiple materials

Megan Maneval — Thu, 16 Oct 2025 20:06:43 +0000

Open-source software allows for efficient 3D printing with multiple materials Megan Maneval Thu, 10/16/2025 - 14:06

College of Engineering and Applied Science

Assistant Professor Robert MacCurdy and doctoral student Charles Wade have created an open-source software package that uses functions and code to map not just shapes but where different materials belong in a 3D object. The project has the potential to transform 3D printing by enabling engineers to design multi-material objects more smartly and efficiently.

Traditional

White

Nobel Laureate John Martinis once served as graduate student mentor at ̽��Ƶ

Elizabeth Lock — Thu, 09 Oct 2025 21:04:56 +0000

Nobel Laureate John Martinis once served as graduate student mentor at ̽��Ƶ Elizabeth Lock Thu, 10/09/2025 - 15:04

Like many rockstar scientists, 2025 Nobel Prize in Physics winner John Martinis spent time in Boulder's rich scientific ecosystem. Martinis mentored graduate students and inspired others in quantum computing.

Traditional

White

Student helps uncover new treatment for respiratory syndrome

Megan Maneval — Tue, 30 Sep 2025 17:50:20 +0000

Student helps uncover new treatment for respiratory syndrome Megan Maneval Tue, 09/30/2025 - 11:50

The project, like something straight out of a sci-fi movie, combines RNA-based gene therapy with tiny microrobots for drug transport to help treat acute respiratory distress syndrome.

Traditional

White

14er science: Quantum physicists measure whether time moves faster on a mountaintop

Daniel William Strain — Tue, 23 Sep 2025 20:57:04 +0000

14er science: Quantum physicists measure whether time moves faster on a mountaintop Daniel William… Tue, 09/23/2025 - 14:57

Daniel Strain

Researchers from Colorado have brought a quantum device known as an optical atomic clock to the summit of Colorado's Mount Blue Sky. Their work could, one day, help people navigate without GPS or even predict when a volcano is about to erupt.

Traditional

White

Power electronics researchers awarded $1.5M to advance energy technologies

Megan Maneval — Thu, 18 Sep 2025 20:34:02 +0000

Power electronics researchers awarded $1.5M to advance energy technologies Megan Maneval Thu, 09/18/2025 - 14:34

College of Engineering and Applied Science

Associate Professor Luca Corradini is embarking on a power electronics project, thanks to a $1.5 million award from the U.S. Department of Energy's Advanced Research Projects Agency—Energy.

Traditional

White

It's not a glitch in the matrix, it's a mathematical phenomenon

Megan Maneval — Wed, 17 Sep 2025 13:26:33 +0000

It's not a glitch in the matrix, it's a mathematical phenomenon Megan Maneval Wed, 09/17/2025 - 07:26

Colorado Arts and Sciences Magazine

̽��Ƶ applied mathematician Mark Hoefer and colleagues answer a longstanding question of how to understand tidal bores in multiple dimensions.

Traditional

White

Construction secrets of honeybees: Study reveals how bees build hives in tricky spots

Amber Elise Carlson — Thu, 11 Sep 2025 11:07:20 +0000

Construction secrets of honeybees: Study reveals how bees build hives in tricky spots Amber Elise Carlson Thu, 09/11/2025 - 05:07

Amber Carlson , Nicholas Goda

From left, Francisco López Jiménez, Orit Peleg and graduate student Richard Terrile inspect the honeycomb in a bee hive. (Credit: Patrick Campbell)

On a hot summer day in Colorado, European honeybees (Apis mellifera L.) buzz around a cluster of hives near Boulder Creek. Worker bees taking off in search of water, nectar and pollen mingle with bees that have just returned from the field. Inside the hives, walls of hexagons are beginning to take shape as the bees build their nests.

“Building a hive is a beautiful example of honeybees solving a problem collectively,” said Orit Peleg, associate professor in ̽��Ƶ’s Department of Computer Science. “Each bee has a little bit of wax, and each bee knows where to deposit it, but we know very little about how they make these decisions.”

In an August 2025 study in PLOS Biology, Peleg’s research group collaborated with Francisco López Jiménez, associate professor in CU’s Ann and H.J. Smead Department of Aerospace Engineering Sciences, and his group to offer new insight into how bees work their hive-making magic—even in the most challenging of building sites.

The new findings could spark ideas for new bio-inspired structures or even new ways to approach 3D printing.

How and why bees build honeycomb

Honeybees can build nests in any number of places, whether it’s a manmade box, a hole in a tree trunk or an empty space inside someone’s attic. When a bee colony finds somewhere new to call home, the bees build their hive out of honeycomb—a waxy structure filled with hexagonal cells—on whatever surfaces are around.

Building a beehive is hard work, and it consumes a lot of resources. It all starts with honey, the nutrient-dense superfood that helps bee colonies survive the winter.

To make honey, bees spend the warmest months gathering nectar from flowers. The nectar mixes with enzymes in the bees’ saliva, and the bees store it in honeycomb cells until it dries and thickens.

It takes roughly 2 million visits to flowers for bees to gather enough nectar to make a pound of honey. Then, each worker bee must eat about 8 ounces of honey to produce a single ounce of the wax they need to build more honeycomb.

If the surface of their building site is irregular, the bees have to expend even more resources building it, and the resulting comb can be harder to use. So efficiency is key.

In an ideal world, bees try to build honeycomb with nearly perfect hexagonal cells that they use for storing food and raising young larvae into adults. Mathematically, the hexagonal shape is ideal for using as little wax as possible to create as much storage space as possible in each cell.

The honeycomb cells are usually a consistent size, but when bees are forced to build comb on odd surfaces, they start making irregular cells that take more wax to build and aren’t as optimal for storage or brood rearing.

Irregular surfaces: A puzzle for bees to solve

This hive frame shows a foundation with a smaller cell size than what bees would typically build. The bees adjusted their building strategies to adapt. (Credit: Patrick Campbell)

Golnar Gharooni Fard, the lead author of the new study and a former CU graduate student, said her main goal in the study was to understand how bees work together to solve the structural problems they might run into.

“We wanted to find the rules of decision-making in a distributed colony,” Fard said.

The researchers 3D printed panels, or foundations, for bees to build comb on. The team imprinted the foundations with shallow hexagonal patterns with differing cell sizes—some larger, some smaller, and some closer to an average cell size—and added the foundations to hives for the bees to use.

Next, the researchers used X-ray microscopy to analyze patterns in the comb the bees built on each type of foundation. Depending on which foundation they were given, the bees used strategies like merging cells together, tilting the cells at an angle or layering them on top of one another to build usable honeycomb.

Giving bees these different surfaces to work with was like giving them puzzles they had to solve, said López Jiménez.

“All those things happen in nature. If they're building honeycomb on a tree, and at some point they get to the end of the branch, the branch might not be super flat, and they need to figure that out,” he said.

It’s still not clear why bees use the strategies they use in all situations. That’s a question the researchers hope to continue exploring.

Meanwhile, the team sees numerous possible applications for their findings. For example, honeycomb could inspire designs for efficient, lightweight structures such as those used in aerospace engineering.

López Jiménez also likened the honeycomb building process to 3D printing, where each bee gradually adds tiny bits of wax to the larger structure.

“The bees take turns, and they organize themselves, and we don't know how that happens,” he said. “Can we learn from how the bees organize labor or how they distribute themselves?”

CU graduate student Chethan Kavaraganahalli Prasanna was also part of the research team.

Beyond the Story

Our research impact by the numbers:

$742 million in research funding earned in 2023–24
No. 5 U.S. university for startup creation
$1.4 billion impact of ̽��Ƶ's research activities on the Colorado economy in 2023–24

Follow ̽��Ƶ on LinkedIn

In a new study, CU researchers found that honeybees used adaptive strategies to build stable, usable honeycomb on irregular and imperfect surfaces.

Traditional

Bees move about their hive on a summer day. (Credit: Patrick Campbell)

White

New optical technique could transform brain imaging in animals

Megan Maneval — Tue, 09 Sep 2025 17:02:58 +0000

New optical technique could transform brain imaging in animals Megan Maneval Tue, 09/09/2025 - 11:02

College of Engineering and Applied Science

̽��Ƶ postdoc Catherine Saladrigas is helping bring high-resolution imaging into miniature microscopes for neuroscience research. The research group tackled how to miniaturize complex optical systems without sacrificing resolution or contrast.

Traditional

White

Physicists have created a new 'time crystal'—it won't power a time machine but could have many other uses

Daniel William Strain — Fri, 05 Sep 2025 14:22:52 +0000

Physicists have created a new 'time crystal'—it won't power a time machine but could have many other uses Daniel William… Fri, 09/05/2025 - 08:22

Daniel Strain

Imagine a clock that doesn’t have electricity, but its hands and gears spin on their own for all eternity.

In a new study, physicists at ̽��Ƶ have used liquid crystals, the same materials that are in your phone display, to create such a clock—or, at least, as close as humans can get to that idea. The team’s advancement is a new example of a “time crystal.” That’s the name for a curious phase of matter in which the pieces, such as atoms or other particles, exist in constant motion.

A time crystal as seen under a microscope. (Credit: Zhao & Smalyukh, 2025, Nature Materials; CC image: https://creativecommons.org/licenses/by-nc-nd/4.0/)

The researchers aren’t the first to make a time crystal, but their creation is the first that humans can actually see, which could open a host of technological applications.

“They can be observed directly under a microscope and even, under special conditions, by the naked eye,” said Hanqing Zhao, lead author of the study and a graduate student in the Department of Physics at ̽��Ƶ.

He and Ivan Smalyukh, professor of physics and fellow with the Renewable and Sustainable Energy Institute (RASEI), published their findings Sept. 4 in the journal "Nature Materials."

In the study, the researchers designed glass cells filled with liquid crystals—in this case, rod-shaped molecules that behave a little like a solid and a little like a liquid. Under special circumstances, if you shine a light on them, the liquid crystals will begin to swirl and move, following patterns that repeat over time.

Under a microscope, these liquid crystal samples resemble psychedelic tiger stripes, and they can keep moving for hours—similar to that eternally spinning clock.

“Everything is born out of nothing,” Smalyukh said. “All you do is shine a light, and this whole world of time crystals emerges.”

Zhao and Smalyukh are members of the Colorado satellite of the International Institute for Sustainability with Knotted Chiral Meta Matter (WPI-SKCM2) with headquarters at Hiroshima University in Japan, an international institute with missions to create artificial forms of matter and contribute to sustainability.

Video of a time crystal in motion. (Credit: Smalyukh Lab)

A computer simulation reveals the inner workings of a time crystal. A beam of light, blue arrow, causes dye molecules, red rods, to change their orientation, driving motion in liquid crystals below. (Credit: Smalyukh Lab)

By stacking several time crystals on top of each other, physicists can create more complex patterns, including what they refer to as a "time barcode." (Credit: Smalyukh Lab)

Crystals in space and time

Time crystals may sound like something out of science fiction, but they take their inspiration from naturally occurring crystals, such as diamonds or table salt.

Nobel laureate Frank Wilczek first proposed the idea of time crystals in 2012. You can think of traditional crystals as “space crystals.” The carbon atoms that make up a diamond, for example, form a lattice pattern in space that is very hard to break apart. Wilczek wondered if it would be possible to build a crystal that was similarly well organized, except in time rather than space. Even in their resting state, the atoms in such a state wouldn’t form a lattice pattern, but would move or transform in a never-ending cycle—like a GIF that loops forever.

Wilczek’s original concept proved impossible to make, but, in the years since, scientists have created phases of matter that get reasonably close.

In 2021, for example, physicists used Google’s Sycamore quantum computer to create a special network of atoms. When the team gave those atoms a flick with a laser beam, they underwent fluctuations that repeated multiple times.

Dancing crystals

In the new study, Zhao and Smalyukh set out to see if they could achieve a similar feat with liquid crystals.

Smalyukh explained that if you squeeze on these molecules in the right way, they will bunch together so tightly that they form kinks. Remarkably, these kinks move around and can even, under certain conditions, behave like atoms.

“You have these twists, and you can’t easily remove them,” Smalyukh said. “They behave like particles and start interacting with each other.”

In the current study, Smalyukh and Zhao sandwiched a solution of liquid crystals in between two pieces of glass that were coated with dye molecules. On their own, these samples mostly sat still. But when the group hit them with a certain kind of light, the dye molecules changed their orientation and squeezed the liquid crystals. In the process, thousands of new kinks suddenly formed.

Those kinks also began interacting with each other following an incredibly complex series of steps. Think of a room filled with dancers in a Jane Austen novel. Pairs break apart, spin around the room, come back together, and do it all over again. The patterns in time were also unusually hard to break—the researchers could raise or lower the temperature of their samples without disrupting the movement of the liquid crystals.

“That’s the beauty of this time crystal,” Smalyukh said. “You just create some conditions that aren’t that special. You shine a light, and the whole thing happens.”

Zhao and Smalyukh say that such time crystals could have several uses. Governments could, for example, add these materials to bills to make them harder to counterfeit—if you want to know if that $100 bill is genuine, just shine a light on the “time watermark” and watch the pattern that appears. By stacking several different time crystals, the group can create even more complicated patterns, which could potentially allow engineers to store vast amounts of digital data.

“We don’t want to put a limit on the applications right now,” Smalyukh said. “I think there are opportunities to push this technology in all sorts of directions.”

A team at ̽��Ƶ has made a curious state of matter in which particles move constantly—like a clock with hands and gears that spin forever, even without electricity to keep them going.

Traditional

The stripes in a time crystal seen under the microscope. (Credit: Zhao & Smalyukh, 2025, Nature Materials; CC image: https://creativecommons.org/licenses/by-nc-nd/4.0/)

White

The stripes in a time crystal seen under a microscope. (Credit: Zhao & Smalyukh, 2025, Nature Materials; CC image: https://creativecommons.org/licenses/by-nc-nd/4.0/)

Science & Technology

5 ways to make AI more trustworthy

How trust forms

1. It knows its users.

2. It’s reliable, ethical and transparent.

3. It takes context into account.

4. It’s easy to use and asks users how it’s doing.

5. When trust is lost, it adapts to rebuild it.

Related Articles

Open-source software allows for efficient 3D printing with multiple materials

Related Articles

Nobel Laureate John Martinis once served as graduate student mentor at ̽��Ƶ

Related Articles

Student helps uncover new treatment for respiratory syndrome

Related Articles

14er science: Quantum physicists measure whether time moves faster on a mountaintop

Related Articles

Power electronics researchers awarded $1.5M to advance energy technologies

Related Articles

It's not a glitch in the matrix, it's a mathematical phenomenon

Related Articles

Construction secrets of honeybees: Study reveals how bees build hives in tricky spots

How and why bees build honeycomb

Irregular surfaces: A puzzle for bees to solve

Related Articles

New optical technique could transform brain imaging in animals

Related Articles

Physicists have created a new 'time crystal'—it won't power a time machine but could have many other uses

Crystals in space and time

Dancing crystals

Related Articles

Science & Technology

5 ways to make AI more trustworthy

How trust forms

1. It knows its users.

2. It’s reliable, ethical and transparent.

3. It takes context into account.

4. It’s easy to use and asks users how it’s doing.

5. When trust is lost, it adapts to rebuild it.

Related Articles

Open-source software allows for efficient 3D printing with multiple materials

Related Articles

Nobel Laureate John Martinis once served as graduate student mentor at ̽����Ƶ

Related Articles

Student helps uncover new treatment for respiratory syndrome

Related Articles

14er science: Quantum physicists measure whether time moves faster on a mountaintop

Related Articles

Power electronics researchers awarded $1.5M to advance energy technologies

Related Articles

It's not a glitch in the matrix, it's a mathematical phenomenon

Related Articles

Construction secrets of honeybees: Study reveals how bees build hives in tricky spots

How and why bees build honeycomb

Irregular surfaces: A puzzle for bees to solve

Related Articles

New optical technique could transform brain imaging in animals

Related Articles

Physicists have created a new 'time crystal'—it won't power a time machine but could have many other uses

Crystals in space and time

Dancing crystals

Related Articles

Nobel Laureate John Martinis once served as graduate student mentor at ̽��Ƶ