United States – Page 44

April 24, 2026

Three reasons why DeepSeek’s new model matters

On Friday, Chinese AI firm DeepSeek released a preview of V4, its long-awaited new flagship model. Notably, the model can process much longer prompts than its last generation, thanks to a new design that helps it handle large amounts of text more efficiently. Like DeepSeek’s previous models, V4 is open source, meaning it is available for anyone to download, use, and modify.

V4 marks DeepSeek’s most significant release since R1, the reasoning model it launched in January 2025. R1, which was trained on limited computing resources, stunned the global AI industry with its strong performance and efficiency, turning DeepSeek from a little-known research team into China’s best-known AI company almost overnight. It also helped set off a wave of open-weight model releases from other Chinese AI firms.

DeepSeek has kept a relatively low profile since then—but earlier this month, it effectively teased V4’s release when it added “expert” and “flash” modes to the online version of its model, prompting speculation that the updates were tied to a bigger upcoming release.

While the company has become a powerful symbol of China’s AI ambitions, its big return to cutting-edge frontier models comes after months of scrutiny—including major personnel departures, delays to previous model launches, and growing scrutiny from both the US and Chinese governments.

So, will V4 shake the AI field the way R1 did? Almost certainly not, but here are three big reasons why this release matters.

1. It breaks new ground for an open-source model.

As with R1 before it, DeepSeek claims that V4’s performance rivals the best models available at a fraction of the price. This is great news for developers and for companies using the tech, because it means they can access frontier AI capabilities on their own terms, and without worrying about skyrocketing costs.

The new model comes in two versions, both of which are available on DeepSeek’s website and in its app, with API access also open to developers. V4-Pro is a larger model built for coding and complex agent tasks, and V4-Flash is a smaller version designed to be faster and cheaper to run. Both versions offer reasoning modes, in which the model can carefully parse a user’s prompt and show each step as it works through the problem.

For V4-Pro, DeepSeek charges $1.74 per million input tokens and $3.48 per million output tokens, a fraction of the cost of comparable models from OpenAI and Anthropic. V4-Flash is even cheaper, at about $0.14 per million input tokens and about $0.28 per million output tokens, making it one of the cheapest top-tier models available. This would make it a very appealing model to build applications on.

In terms of performance, V4 is, perhaps unsurprisingly, a huge jump from R1—and it seems to be a strong alternative to just about all the latest big AI models. On the major benchmarks, according to results shared by the company, DeepSeek V4-Pro competes with leading closed-source models, matching the performance of Anthropic’s Claude-Opus-4.6, OpenAI’s GPT-5.4, and Google’s Gemini-3.1. And compared to other open-source models, such as Alibaba’s Qwen-3.5 or Z.ai’s GLM-5.1, DeepSeek V4 exceeds them all on coding, math, and STEM problems, making it one of the strongest open-source models ever released.

DeepSeek also says that V4-Pro now ranks among the strongest open-source models on benchmarks for agentic coding tasks and performs well on other tests that measure ability to carry out multistep problems. Its writing ability and world knowledge also leads the field, according to benchmarking results shared by the company.

In a technical report released alongside the model, DeepSeek shared results from an internal survey of 85 experienced developers: More than 90% included V4-Pro among their top model choices for coding tasks.

DeepSeek says it has specifically optimized V4 for popular agent frameworks such as Claude Code, OpenClaw, and CodeBuddy.

2. It delivers on a new approach to memory efficiency.

One of the key innovations of V4 is its long context window—the amount of text the model can process at once. Both versions can handle 1 million tokens, which is large enough to fit all three volumes of The Lord of the Rings and The Hobbit combined. The company says this context window size is now the default across all DeepSeek services and it matches what is offered by cutting-edge versions of models like Gemini and Claude.

But it’s important to know not just that DeepSeek has made this leap, but how it did so. V4 makes significant architectural changes to the company’s former models—especially in the attention mechanism, which is the feature of AI models that helps them understand each part of a prompt in relation to the rest. As the prompt text gets longer, these comparisons become much more costly, making attention one of the main bottlenecks for long-context models.

DeepSeek’s innovation was to make the model more selective about what it pays attention to. Instead of treating all earlier text as equally important, V4 compresses older information and focuses on the parts most likely to matter in the present moment, while still keeping nearby text in full so it does not miss important details.

DeepSeek says this sharply reduces the cost of using long context. In a 1-million-token context, V4-Pro uses only 27% of the computing power required by its previous model, V3.2, while cutting memory use to 10%. The reduction in V4-Flash is even larger, using just 10% of the computing power and 7% of the memory. In practice, this could make it cheaper to build tools that need to work across huge amounts of material, such as an AI coding assistant that can read an entire codebase or a research agent that can analyze a long archive of documents without constantly forgetting what came before.

DeepSeek’s interest in long context windows didn’t start with V4. Over the past year and a half, the company has quietly published a series of papers on how AI models “remember” information, experimenting with compression and mathematical techniques to extend what AI models could realistically handle.

3. It marks the first steps on the hard road away from Nvidia.

V4 is DeepSeek’s first model optimized for domestic Chinese chips, such as Huawei’s Ascend—a move that has turned the launch into something of a test of whether China’s homegrown AI industry can begin to loosen its dependence on US chip giant Nvidia.

This was largely expected, since The Information reported earlier this month that DeepSeek did not give American chipmakers like Nvidia and AMD early access to V4, though prerelease access is common to allow chipmakers to optimize support of the new model ahead of a launch. Instead, the company reportedly gave early access only to Chinese chipmakers.

On Friday, Huawei said its Ascend supernode products, based on the Ascend 950 series, would support DeepSeek V4. This means that companies and individuals who want to run their own modified version of Deepseek V4 will be able to use Huawei chips easily.

Reuters previously reported that Chinese government officials recommended that DeepSeek integrate Huawei chips in its training process. And this pressure fits a broader pattern in China’s industrial policy: Strategic sectors are often pushed, and sometimes effectively required, to align with national self-reliance goals. But there’s a particular urgency when it comes to AI. Since 2022, US export controls have cut Chinese firms off from Nvidia’s most powerful chips, and they later also restricted access to downgraded China-market versions. Beijing’s response has been to accelerate the push for a domestic AI stack, from chips to software frameworks to data centers.

Chinese authorities have reportedly been pushing data centers and public computing projects to use more domestic chips, including through reported bans on foreign-made chips, sourcing quotas, and requirements to pair Nvidia chips with Chinese alternatives from companies such as Huawei and Cambricon.

Still, replacing Nvidia is not as simple as swapping one chip for another. Nvidia’s advantage lies not only in its chips, but in the software ecosystem developers have spent years building around them. Moving to Huawei’s Ascend chips means adapting model code, rebuilding tools, and proving that systems built around those chips are stable enough for serious use.

To be clear, DeepSeek does not appear to have fully moved beyond Nvidia. The company’s technical report reveals that it is using Chinese chips to run the model for inference, or when someone asks the model to complete a task. But Liu Zhiyuan, a computer science professor at Tsinghua University, told MIT Technology Review that DeepSeek appears to have adapted only part of V4’s training process for Chinese chips. The report does not say whether some key long-context features were adapted to domestic chips, so Liu says V4 may still have been trained mainly on Nvidia chips. Multiple sources who spoke on the condition of anonymity, due to political sensitivity around these issues, told MIT Technology Review that Chinese chips still don’t perform as well as Nvidia chips but are better suited for inference than training.

DeepSeek is also tying the future costs of V4 to this hardware shift. The company says V4-Pro prices could fall significantly after Huawei’s Ascend 950 supernodes begin shipping at scale in the second half of this year.

If that works, V4 could be an early sign that China is successfully building a parallel AI infrastructure.

April 24, 2026

One Biosciences Chooses Albany, NY, as Its U.S. Location

Paris-based One Biosciences, an Institut Curie-backed startup, plans to set up, staff, and equip a high-complexity lab and computational analytics operation in Albany, NY, as its first U.S. location.

Empire State Development is supporting this expansion with up to $525,000 in performance-based Excelsior Jobs Program tax credits in exchange for the company’s job commitments, which anticipate 42 life science jobs and $18 million in investments over the next five years.

Officials at One Biosciences say the company will bring its proprietary technology to the first-of-its-kind hub in Albany to address the unmet clinical and scientific needs to characterize the tumor ecosystem by means of a single-cell profiling approach.

“We are excited to accelerate support of our pharma, biotech, and academic collaborators through our AI-driven single-cell technologies, which will ultimately benefit physicians and their patients,” added Vincent Miller, MD, executive chairman, One Biosciences. “The local Albany life sciences ecosystem gives us access to a community of like-minded researchers and physicians committed to leveraging technology to improve health and is an ideal location from where to serve the U.S. globally.”

“Life science research and development is vital to creating the treatments that help people heal, survive and live longer,” said New York governor Kathy Hochul. “Through our targeted efforts, we are working to ensure that cutting edge companies like One Biosciences not only grow here, but that the next generation of medical breakthroughs happen in New York State.”

The post One Biosciences Chooses Albany, NY, as Its U.S. Location appeared first on GEN – Genetic Engineering and Biotechnology News.

April 24, 2026

AI Learns to Predict Breast Cancer Risk from How Single Cells Respond to Pressure

A study headed by researchers at City of Hope and the University of California, Berkeley has found that physical and mechanical properties of normal human mammary epithelial cells can offer a “functional readout” of biological age and breast cancer susceptibility.

The team created a novel, high-throughput microfluidic platform that can assess women’s breast cancer risk at the cellular level. The mechano-node-pore sensing (mechano-NPS) platform, which the researchers claim is the first of its kind, squeezes individual breast epithelial cells, creating a taxing environment to measure how they deform, recover, and behave under stress.

Using the platform the researchers uncovered an unexpected insight, which is that breast cells appear to have a “mechanical age” separate from a person’s chronological age, demonstrated by how the cells physically respond to stress. For their study the team developed a machine learning classifier, MechanoAge, to estimate chronological age based on the mechanical phenotypes, and a biological age-based risk index, Mechano-RISQ.

“We learned that the older the mechanical age, as determined by how cells respond to being squeezed through our microfluidic device, the higher the risk for breast cancer,” explained Lydia Sohn, PhD, the Almy C. Maynard and Agnes Offield Maynard Chair in Mechanical Engineering at UC Berkeley. The researchers suggest that, as more than 90% of women lack a known genetic predisposition to or a family history of breast cancer, their innovative approach could fill a critical gap in risk assessment and save countless lives.

Sohn is co-senior author of the team’s published paper in eBioMedicine, titled “MechanoAge, a machine learning platform to identify individuals susceptible to breast cancer based on mechanical properties of single cells,” in which they concluded, “Age-related biomechanical changes may represent a fundamental hallmark of cellular function, with distinct mechanical phenotypes underlying critical processes in aging, cancer, and potentially other diseases. Recognizing and utilizing these biomechanical markers could greatly enhance early detection, refine risk stratification, and improve targeted intervention strategies.”

Breast cancer is one of the most frequently diagnosed cancers worldwide and a leading cause of cancer-related mortality among women, the authors noted, and “… has long been the subject of efforts to improve risk stratification and early-detection strategies.”

About 6% of women who develop breast cancer carry known genetic mutations. But for women outside this group, risk is estimated indirectly based on population models or measurements like breast density. These approaches can both overestimate and underestimate women’s individual breast cancer risk, leading to over-screening, under-screening, unnecessary worry or missed warning signs. And despite significant progress in screening technologies and therapeutic interventions, accurately determining which individuals—particularly among those considered average risk—are most likely to develop breast cancer remains what the team calls “one of the most persistent challenges in oncology and public health.”

For these “ostensibly average-risk individuals,” the team added, “it remains difficult to identify those with latent risk that stems from cellular, molecular, and biophysical alterations that current models are not designed to capture.”

Researchers Mark LaBarge of City of Hope (right) and Lydia Sohn (left) UC Berkeley [City of Hope and UC Berkeley]

Currently, there is no non-genetic test available that can identify women at higher risk for breast cancer. A downside to screening mammograms is that they can catch cancer only once it has begun to grow. Co-senior author, Mark LaBarge, PhD, a professor in the Department of Population Sciences at City of Hope, said “For women with a known genetic risk factor for breast cancer, there are things you can do like follow a higher-risk screening protocol. For everybody else, you’re left wondering, ‘Am I at high risk?’”

Emerging evidence links cellular aging and biophysical alterations with cancer susceptibility. For their reported study the researchers used the mechano-NPS platform to profile primary human mammary epithelial cells (HMECs) from women of different ages and risk backgrounds. They also developed a machine learning algorithm that identifies and measures cells that show signs of accelerated aging, quantifying an individual breast cancer risk score.

In this type of mechano-node-pore sensing, an electrical current is measured across a liquid-filled channel, much like how current is measured across a wire. As cells pass through, they disrupt the current, generating measurements about the cells’ size and shape. By making parts of the channel very narrow, researchers squeeze cells, then measure how long it takes each cell to recover its normal shape.

Machine-learning algorithms developed by the researchers were then used to detect differences in cells from older and younger women. The researchers found that the physical properties of breast cells changed with age; cells from older women were stiffer and took longer to bounce back after being squeezed.

Then came a surprising finding: a subset of younger women had cells that behaved like they came from older women. These cells came from women with genetic mutations that put them at high risk of breast cancer. Researchers then refined the algorithm to assign a risk score based on all the mechanical and physical properties measured in the cells. This algorithm successfully identified women with known genetic risks. Next the team used it to compare cells from healthy women, women who had family history of breast cancer and cells taken from the healthy breast of women with breast cancer in the other breast. “Normal epithelial cells from women with germline mutations, strong family history of cancer, or contralateral breast cancer exhibit mechanically aged phenotypes despite normal histology,” the investigators stated. “Together with prior molecular and epigenetic studies, these findings support a model in which accelerated biological aging of mammary epithelia may underpin breast cancer susceptibility across genetic and non-genetic risk groups.”

Using the MechanoAge platform, researchers shifted the scientific lens to the cellular level, calculating risk by looking for physical changes in individual cells. “Mechanical phenotyping captures an integrative cellular state that reflects underlying molecular networks rather than single biomarkers,” the team noted. “Mechano-RISQ offers a proof of principle approach for identifying individuals at elevated risk of breast cancer, especially among average-risk populations, and may complement existing risk models by incorporating biophysical measures of mammary epithelial cell aging.”

“With accuracy, we were able to figure out which women were at high risk of breast cancer and which women didn’t seem to be,” LaBarge said. “By translating physical changes in cells into quantifiable data, this tool gives women something tangible to discuss with their doctors—not just risk estimates, but evidence drawn directly from their own cells.” In their paper the scientists further stated, “This approach could enable earlier, individualized risk stratification, particularly for women who lack identifiable high-risk mutations yet harbor susceptible tissue states.”

Importantly, the AI platform uses simple electronics that would be easy and affordable to replicate on a large scale. “Our team isn’t the first to measure the mechanical properties of cells; however, other approaches require advanced imaging technology that’s expensive, cumbersome and has limited availability,” said Sohn. “In contrast, MechanoAge uses computer chips that are simpler than an Apple Watch and ‘Radio Shack parts’ that are cheap and easy to assemble, potentially making the device highly scalable.”

While engineers study the aging of materials such as metals, concrete and polymers, this is the first time that mechanical age has been quantified in biological cells. The finding that cells have a “mechanical age” separate from the individual’s chronological age would not have been possible without MechanoAge.

This work grew out of more than 12 years of collaboration between the two labs, combining engineering innovation with cancer and aging biology. The long-term partnership enabled discoveries that neither group could have reached alone. “It’s a true collaboration. We’ve learned a lot from each other,” Sohn said. “In my view, this is what happens when you have a real collaboration that develops over a long time,” LaBarge added. “This result is not what we imagined at the beginning.”

The post AI Learns to Predict Breast Cancer Risk from How Single Cells Respond to Pressure appeared first on GEN – Genetic Engineering and Biotechnology News.

April 24, 2026

AI in Oncology Takes Off, Tackling HIV and Liver Disease, Pharma’s Recent Gains

Some GEN editors were in sunny San Diego covering the hottest research, trends, and products from the American Association for Cancer Research meeting. We kick things off with news from America’s Finest City, particularly around the growing role of AI in oncology. Then we dive into two new research studies. In the first, scientists used CRISPR to identify genes in primary CD4+ T cells that promote or restrict HIV infection. The second study described engineered implantable liver constructs that could eventually serve as a stopgap for patients waiting for donor transplants. Finally, the acquisitions keep coming as Eli Lilly scoops up CAR T cell therapy developer Kelonia for $7B. Also, Revolution Medicines has shared some impressive data from a Phase III trial of its pancreatic cancer drug.

Listed below are links to the GEN stories referenced in this episode of Touching Base:

AACR 2026: A Video Update from San Diego
By Julianna LeMieux, PhD, and Damian Doherty, GEN, April 21, 2026

AACR 2026 Video Update: Cancer Research Edges Toward an AI-Driven Era
By Fay Lin, PhD, and Jonathan Grinstein, PhD, GEN, April 22, 2026

Using AI in Healthcare Ethically by Considering Humanity
By Corinna Singleman, PhD, IPM, November 18, 2025

10x Genomics Unveils Atera Spatial Platform at AACR Meeting
By Julianna LeMieux, PhD, GEN, April 19, 2026

CRISPR Screens Map Human T‑Cell Genes That Promote or Block HIV Infection
GEN, April 20, 2026

Synthetic Biology and Tissue Engineering Grow Liver Tissue In‑Body
GEN, April 20, 2026

StockWatch: Revolution’s Phase III Pancreatic Cancer Data Dazzles Investors, Analysts
By Alex Philippidis, GEN Edge, April 19, 2026

Lilly to Acquire Kelonia for Up to $7B, Expanding Cancer Cell Therapy Pipeline
By Alex Philippidis, GEN Edge, April 20, 2026

Touching Base Podcast
Hosted by Corinna Singleman, PhD

Behind the Breakthroughs
Hosted by Jonathan D. Grinstein, PhD

The post AI in Oncology Takes Off, Tackling HIV and Liver Disease, Pharma’s Recent Gains appeared first on GEN – Genetic Engineering and Biotechnology News.

April 24, 2026

The Download: supercharged scams and studying AI healthcare

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

We’re in a new era of AI-driven scams

When ChatGPT was released in late 2022, it showed how easily generative AI could create human-like text. This quickly caught the eye of cybercriminals, who began using LLMs to compose malicious emails. Since then, they’ve adopted AI for everything from turbocharged phishing and hyperrealistic deepfakes to automated vulnerability scans.

Many organizations are now struggling to cope with the sheer volume of cyberattacks. AI is making them faster, cheaper, and easier to carry out, a problem set to worsen as more cybercriminals adopt these tools—and their capabilities improve. Read the full story on how AI is reshaping cybercrime.

—Rhiannon Williams

“Supercharged scams” is one of the 10 Things That Matter in AI Right Now, our essential guide to what’s really worth your attention in the field.

Subscribers can watch an exclusive roundtable unveiling the technologies and trends on the list, with analysis from MIT Technology Review’s AI reporter Grace Huckins and executive editors Amy Nordrum and Niall Firth.

Healthcare AI is here. We don’t know if it actually helps patients.

Doctors are using AI to help them with notetaking. AI-based tools are trawling through patient records, flagging people who may require certain support or treatments. They are also used to interpret medical exam results and X-rays.

A growing number of studies suggest that many of these tools can deliver accurate results. But there’s a bigger question here: Does using them actually translate into better health outcomes for patients? We don’t yet have a good answer—here’s why.

—Jessica Hamzelou

The story is from The Checkup, our weekly newsletter that gives you the latest from the worlds of health and biotech. Sign up to receive it in your inbox every Thursday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 DeepSeek has unveiled its long-awaited new AI model
The Chinese company has just launched preview versions of DeepSeek-V4. (CNN)
+It says V4 is the most powerful open-source platform. (Bloomberg $)
+ And rivals top closed-source models from OpenAI and DeepMind. (SCMP)
+ The model is adapted for Huawei chip technology. (Reuters $)

2 More countries are curbing children’s social media access
Norway is set to enforce the latest ban. (Reuters $)
+ The Philippines could follow soon. (Bloomberg $)
+ Americans are pushing to get AI out of schools. (The New Yorker)

3 The US has accused China of mass AI theft as tensions rise
A White House memo claims Chinese firms are exploiting American models. (BBC)
+ Beijing calls the accusations “slander.” (Ars Technica)

4 OpenAI set itself apart from Anthropic by widely releasing its new model
It’s releasing GPT-5.5 to all ChatGPT users, despite cybersecurity concerns. (NYT $)
+ OpenAI says the new model is better at coding and more efficient. (The Verge)

5 Meta is cutting 10% of jobs to offset AI spending
Roughly 8,000 layoffs are set to be announced on May 20. (QZ)
+ Anti-AI protests are growing. (MIT Technology Review)

6 Palantir is facing a backlash from employees
Thanks to its work with ICE and the Trump administration. (Wired $)
+ Surveillance tech is reshaping the fight for privacy. (MIT Technology Review)

7 The era of free access to advanced AI is coming to an end
AI labs are under mounting pressure to start turning profits. (The Verge)

8 Elon Musk’s feud with Sam Altman is heading to court
The case has already revealed several unflattering secrets. (WP $)

9 A new movement is encouraging people to ditch their smartphones for a month
“Month Offline” is like a Dry January for smartphones. (The Atlantic)

10 Spotify has revealed its most-streamed music of the last 20 years
Featuring Taylor Swift, Bad Bunny, and The Weeknd. (Gizmodo)

Quote of the day

“We want a childhood where children get to be children. Play, friendships, and everyday life must not be taken over by algorithms and screens.”

—Norwegian Prime Minister Jonas Gahr Store announces age restrictions for social media.

One More Thing

The search for extraterrestrial life is targeting Jupiter’s icy moon Europa

As astronomers have discovered more about Europa over the past few decades, Jupiter’s fourth-largest moon has excited planetary scientists interested in the geophysics of alien worlds.

All that water and energy—and hints of elements essential for building organic molecules —point to an extraordinary possibility. In the depths of its ocean, or perhaps crowded in subsurface lakes or below icy surface vents, Jupiter’s big, bright moon could host life.

To find further evidence, NASA is now searching for signs of alien existence on Europa. Read the full story on the mission.

—Stephen Ornes

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line.)

+ Here’s a fun look at the secret collaborations of pop history.
+ Meet the mannequins showing how the “ideal” body has evolved.
+ A photographer has cataloged all 12,795 objects in her home into an archive of a life.
+ Slime molds are unexpectedly beautiful when viewed through these high-detail macro shots.

April 24, 2026

Health-care AI is here. We don’t know if it actually helps patients.

I don’t need to tell you that AI is everywhere.

Or that it is being used, increasingly, in hospitals. Doctors are using AI to help them with notetaking. AI-based tools are trawling through patient records, flagging people who may require certain support or treatments. They are also used to interpret medical exam results and X-rays.

We don’t yet have a good answer.

That’s what Jenna Wiens, a computer scientist at the University of Michigan, and Anna Goldenberg of the University of Toronto, argue in a paper published in the journal Nature Medicine this week.

Wiens tells me she has spent years investigating how AI might benefit health care. For the first decade of her career she tried to pitch the technology to clinicians. Over the last few years, she says, it’s as though “a switch flipped.” Health-care providers not only appear much more interested in the promise of these technologies, they have also begun rapidly deploying them.

The problem is that many providers aren’t rigorously assessing how well they actually work.

Take “ambient AI” tools, for example. Also known as AI scribes, they “listen” to conversations between doctors and patients, then transcribe and summarize them. Multiple tools are available, and they are already being widely adopted by health-care providers.

A few months ago, a staffer at a major New York medical center who develops AI tools for doctors told me that, anecdotally, medics are “overjoyed” by the technology—it allows them to focus all their attention on their patients during appointments, and it saves them from a lot of time-consuming paperwork. Early studies support these anecdotes and suggest that the tools can reduce clinician burnout.

That’s all well and good. But what about patient health outcomes? “[Researchers] have evaluated provider or clinician and patient satisfaction, but not really how these tools are affecting clinical decision-making,” says Wiens. “We just don’t know.”

The same holds true for other AI-based technologies used in health-care settings. Some are used to predict patients’ health trajectories, others to recommend treatments. They are designed to make health care more effective and efficient.

But even a tool that is “accurate” won’t necessarily improve health outcomes. AI might speed up the interpretation of a chest X-ray, for example. But how much will a doctor rely on its analysis? How will that tool affect the way a doctor interacts with patients or recommends treatment? And ultimately: What will this mean for those patients?

The answers to those questions might vary between hospitals or departments and could depend on clinical workflows, says Wiens. They might also differ between doctors at various stages of their careers.

Take the AI scribes, as another example. Some research on AI use in education suggests that such tools can impact the way people cognitively process information. Could they affect the way a doctor processes a patient’s information? Will the tools affect the way medical students think about patient data in a way that impacts care? These questions need to be explored, says Wiens. “We like things that save us time, but we have to think about the unintended consequences of this,” she says.

In a study published in January 2025, Paige Nong at the University of Minnesota and her colleagues found that around 65% of US hospitals used AI-assisted predictive tools. Only two-thirds of those hospitals evaluated their accuracy. Even fewer assessed them for bias.

The number of hospitals using these tools has probably increased since then, says Wiens. Those hospitals, or entities other than the companies developing the tools, need to evaluate how much they help in specific settings. There’s a possibility that they could leave patients worse off, although it’s more likely that AI tools just aren’t as beneficial as health-care providers might assume they are, says Wiens.

“I do believe in the potential of AI to really improve clinical care,” says Wiens, who stresses that she doesn’t want to stop the adoption of AI tools in health care. She just wants more information about how they are affecting people. “I have to believe that in the future it’s not all AI or no AI,” she says. “It’s somewhere in between.”

This article first appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday, and read articles like this first, sign up here.

April 24, 2026

Caring for Dementia Caregivers in Ethnic Immigrant Communities

Conditions: Stress and Well-being of Family Caregivers of Persons With Dementia

Interventions: Behavioral: K-Savvy; Behavioral: Healthy Living

Sponsors: University of Southern California

Not yet recruiting

April 23, 2026

Exploring the Cultural Adaptation of an Ongoing Evidence-Based Intervention for Chinese and Korean American Dementia Caregivers: Descriptive Study

Background: The aging and caregiving population is becoming increasingly diverse in the United States, leading to a growing need for culturally adapted interventions to address the unique needs of underrepresented groups, such as Asian Americans. However, interventions targeting Asian Americans and exploring cultural adaptation strategies remain limited in dementia caregiving research. Objective: This study aimed to describe the cultural adaptation process of an evidence-based intervention for Chinese and Korean American dementia caregivers, called the New York University Caregiver Intervention–Enhanced Support. Methods: We conducted a deductive content analysis and categorized our adaptation strategies into 5 elements: content, context, relationship fidelity and core elements, engagement, and cultural competence. Timing and types of responses to each adaptation strategy were also observed. Two authors conducted the initial analysis, and additional team members finalized the synthesis through discussion. The Template for Intervention Description and Replication (TIDieR) checklist was used to guide the methodological rigor. Results: Twenty-four major adaptations were identified and categorized. For content, we translated materials, used culturally relevant terms, incorporated ethnic-specific surveys and resources, created social media support groups on platforms widely used by the targeted population, and extended the time allocated to complete the 6 counseling sessions. Context adaptation included expanding the range of individuals eligible for family counseling sessions to include fictive kin, using online and social media apps for communication, cultural matching and training of staff, and partnerships with relevant community organizations. Relationship fidelity and core elements involved consulting with community experts, conducting focus group interviews with caregivers, having regular meetings with the developer of the original intervention and an experienced New York University Caregiver Intervention–Enhanced Support clinician as well as experts in Chinese and Korean culture, and continuing regular counseling supervision. To enhance engagement, we provided clear explanations of the study procedure, which emphasized the benefits in participants’ native languages and matched participants with social workers who shared the same cultural backgrounds. We also used a step-by-step contact approach and prolonged communication, explained staff roles to build rapport, and offered participant compensation. Finally, cultural competence was reflected in tailoring counseling techniques with respect for cultural beliefs, the use of euphemistic language for taboo subjects, and culturally appropriate refreshments to show respect and build interpersonal relationships. Conclusions: We systematically adjusted a counseling-based intervention, an approach less familiar among Asian Americans, to fit the cultural characteristics of the target population. A contribution of this study is using an integrated, theory-driven approach that combines 2 cultural adaptation frameworks while also capturing real-time adaptations informed by external feedback and self-reflection. This work provides a practical model for adapting evidence-based interventions to serve Chinese and Korean American dementia caregivers and may inform future adaptations for other East Asian populations. Trial Registration: ClinicalTrial.gov NCT05461495; https://clinicaltrials.gov/study/NCT05461495
<img src="https://jmir-production.s3.us-east-2.amazonaws.com/thumbs/e08f825f332b5b67120fb0f3ebf4f04c" />

April 23, 2026

STAT+: Trump celebrates closing first round of drug pricing deals, promises more ahead

WASHINGTON — President Trump heralded a drug pricing agreement with Regeneron on Thursday, closing the last of 17 deals initially sought by the White House last year.

Regeneron, as part of the private deal, will reduce prices on drugs to Medicaid, provide cholesterol medicine Praluent on TrumpRx for $225, and invest $27 billion in drug development in the United States.

On the same day, Regeneron also announced Food and Drug Administration approval of Otarmeni, the first gene therapy to be greenlit under the agency’s new National Priority Voucher program. In early trials, the drug provided modest hearing gains for people with a rare type of hearing loss, though its development has received pushback from parts of the Deaf community. Regeneron plans to offer the drug at no cost to American patients.

Continue to STAT+ to read the full story…