The Download: coding’s future, the ‘Steroid Olympics,’ and AI-driven science

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.

Anthropic’s Code with Claude showed off coding’s future—whether you like it or not

At Anthropic’s developer event in London this week, Code with Claude, attendees were asked if they’d shipped code written entirely by Claude. Almost half the room raised their hands. Many admitted they hadn’t even read the code before pushing it live.

As tools like Claude Code get better, more and more developers are happy to hand their work off to AI. Anthropic says it wants to push automation as far as it will go. But not everyone is convinced that’s the right approach. 

Read the full story on how AI is reshaping coding for good.

—Will Douglas Heaven

The Enhanced Games fit right in with the rest of 2026’s longevity vibes

This Sunday, 42 athletes will gather in Las Vegas for the inaugural Enhanced Games, a controversial sporting competition that allows the use of performance-enhancing drugs. The goal? To “push the boundaries of human performance.”

The event embodies a zeitgeist of peptide-crazed looksmaxxing, where consumers are encouraged to get thinner than ever, optimize for longevity, and have their “best baby.” In 2026, if you’re not enhancing, what are you even doing?

Find out how the competition reflects our enhancement-obsessed era.

—Jessica Hamzelou

This story is from The Checkup, our weekly newsletter giving you the inside track on all things biotech. Sign up to receive it in your inbox every Thursday.

Google I/O showed how the path for AI-driven science is shifting

—Grace Huckins

During Tuesday’s Google I/O keynote, Demis Hassabis, the CEO of Google DeepMind, proclaimed that we are “standing in the foothills of the singularity.” But what struck me as I listened in the audience was the context in which he said those words.

The contrast reflects two directions for AI in science. One builds specialized systems like WeatherNext for specific problems. The other pushes toward agentic, LLM-based systems that could eventually execute cutting-edge research projects without human involvement.

The big scientific announcement at I/O was Gemini for Science, which leans further into this agent-driven future. It can still call on specialized systems, but Google appears to be transitioning away from them.

Here’s how the shift could affect science.

Can AI learn to understand the world?

Many leading AI researchers have turned their attention to a new kind of system that understands the physical environment: world models. 

Backed by researchers at Google DeepMind, Fei-Fei Li’s World Labs, and Meta’s former Chief AI scientist, Yann LeCun, the idea is gaining serious momentum. Could it change how AI understands reality?

MIT Technology Review editor in chief Mat Honan, senior AI editor Will Douglas Heaven, and AI reporter Grace Huckins unpacked it all in an exclusive Roundtables discussion yesterday.

Subscribers can watch the full recording now.

World models are also one of MIT Technology Review’s 10 Things That Matter in AI Right Now, our list of what’s really worth your attention in the busy, buzzy world of AI.

The must-reads

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

1 Trump has postponed an AI order due to overregulation fears
He said he was concerned it would be “a blocker.” (CNBC)
+ And that he wants to preserve the US’s lead over China in AI. (Reuters $)
+ A source said the delay was because he “just hates regulation.” (Axios)
+ A war over regulation is coming to America. (MIT Technology Review)

2 OpenClaw’s engineers warn that a “vibe-coded slop” crisis is coming
They say AI is flooding the world with bad and even dangerous code. (WSJ $)
+ Now vibe coding is coming to your phone, too. (The Verge)
+ What exactly is vibe coding? (MIT Technology Review)

3 SpaceX has called off the launch of a new Starship prototype
Engineers discovered a ground system glitch. (CNBC)
+ They hope to try again tonight. (Ars Technica)
+ The launch could play a key role in SpaceX’s IPO. (NPR)

4 Meta has settled a school district’s social media addiction lawsuit
It had been sued over the alleged harm caused to students. (BBC)
+ Snap, TikTok, and YouTube have also settled with the district. (NYT $)

5 Bluesky says it’s being hacked by the Kremlin to spread propaganda
It’s fighting Russian efforts to hijack real users’ accounts to post. (NYT $)
+ Now is a good time for doing crime. (MIT Technology Review)

6 Africa’s biggest economies are pushing for AI sovereignty
They aim to reduce their dependence on Big Tech. (Rest of World)
+ New strategies could make Africa a major AI player. (MIT Technology Review)

7 Undersea cables threaten the Gulf’s AI expansion plans
Conflicts have put the fragile critical infrastructure at risk. (Wired $)

8 Waymo is pausing services as robotaxis keep driving into floods
It suspended services in four US cities. (TechCrunch)

9 Microscopic silica spheres may help cool the planet
But some researchers need further convincing. (The Economist $)

10 Spotify will now let subscribers create AI remixes

It’s the first time they can use AI to create content on Spotify. (Guardian)

Quote of the day

“You have AI — actual intelligence.” 

—Apple cofounder Steve Wozniak reassures college graduates about AI’s impact and draws applause, in contrast to the boos received by former Google CEO Eric Schmidt earlier this week, Business Insider reports.

One More Thing

Looking down a neighborhood street where a man in wheelchair has crossed with wife and daughter.

GETTY IMAGES


The future is disabled

Technologies for disability, access, and mobility are often portrayed as objects of empowerment or heroic, life-changing panaceas for social ills. But their benefits are often temporary, lopsided, or reliant on constant investment, care, and attention.

Often, accessibility tech assumes levels of access that don’t exist: reliable internet, smartphones, or affordable devices. Projects frequently overlook the very communities they claim to serve. Yet there’s another way: opening ourselves up to all-access thinking and disabled expertise.

Discover how that approach could create a more livable world for everyone.

—Ashley Shew

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.)

+ Treat your eyes to this magical footage of a lake floating above an ocean.
+ Test your visual recall with this clever game that recreates colors from memory.
+ Take back control of your internet with this dashboard that brings together your favourite social feeds.
+ Peer into the heart of a barred spiral galaxy in this stunning new capture from the James Webb Space Telescope.

Google I/O showed how the path for AI-driven science is shifting

During Tuesday’s Google I/O keynote, Demis Hassabis, the CEO of Google DeepMind, proclaimed that we are currently “standing in the foothills of the singularity.” It was a striking statement—the singularity is the theoretical future moment when AI rapidly exceeds human intelligence and dramatically transforms the world. But what struck me as I listened in the audience was the context in which he said those words. 

He was on stage to close out the session with a segment on scientific AI, the centerpiece of which was a video detailing how the company’s weather prediction software provided an advance alert about Hurricane Melissa’s catastrophic landfall in Jamaica last year—and potentially saved lives. If that software, called WeatherNext, helped anyone escape the storm or better fortify their home, that’s an enormous and meaningful achievement. But it’s hardly evidence of an impending singularity.

The juxtaposition of Hassabis’ lofty rhetoric with the real-world results of WeatherNext highlighted the tension between two very different approaches to AI for science. The first focuses on AI tools, like WeatherNext, that are designed and trained to solve specific scientific problems. The second is agentic, LLM-based systems that could one day execute cutting-edge research projects without human involvement.

This second vision powers a great deal of AI enthusiasm right now, including recent excitement around recursive self-improvement, or the idea that AI systems could eventually become the primary drivers of AI advancement—a process that would get faster and faster as the AI systems grow smarter. And agentic systems are now making real research contributions, sometimes with limited human guidance.

Just this week, Pushmeet Kohli, Google Cloud’s chief scientist, published a piece in a special AI and science issue of the journal Daedalus, writing: “We are moving toward AI that doesn’t just facilitate science but begins to do science.” With autonomous AI scientists on the horizon, it’s harder to justify massive efforts to develop super-specialized tools—even one like AlphaFold, for which DeepMind scientists won a Nobel Prize, or a potentially life-saving system like WeatherNext. It also heralds a far stranger future for science, in which humans and AI systems collaborate as peers—or AI even makes scientific progress on its own.

To be clear, Google does not appear to be abandoning its work on specialized AI for science tools. AlphaGenome and AlphaEarth Foundations, which are trained for genetics and Earth science applications respectively, were released last summer, and the newest version of WeatherNext came out in November.

What’s more, such tools remain extremely popular among scientists. Last year, for instance, Google reported that protein structure predictions from AlphaFold have been used by over three million researchers worldwide. And Isomorphic Labs, a Google subsidiary that aims to use AlphaFold and related technologies to develop new drugs, just raised a $2 billion Series B funding round.

But there are concrete signs of realignment, in both enthusiasm and resources. Last month, the Los Angeles Times reported that Google fellow John Jumper, who won the Nobel for AlphaFold, is now working on AI coding, not on science-specific AI tools. It’s not surprising that Google is assigning its best minds to the coding problem, as the company has recently taken a reputational hit because its coding tools don’t currently stand up to those offered by Anthropic and OpenAI. But it may also signal a prioritization of agentic science on Google’s part, as coding abilities are key to the success of some of those systems. 

Across the industry, agentic researcher systems are showing real potential. This week, OpenAI announced that one of their models had disproved an important mathematics conjecture—perhaps the most meaningful contribution that generative AI has made to mathematics so far, according to some mathematicians.

Importantly, the model used by OpenAI is not specialized for solving mathematical problems, or even for research; according to the company, it’s a general-purpose reasoning model in the vein of GPT-5.5. If general agents can make independent contributions to mathematical research, they might soon be able to do the same in science (though the fact that ideas in science must be verified experimentally makes it a tougher domain for AI).

Google is certainly devoting a lot of attention toward an agent-driven scientific future. The big scientific announcement at I/O was the new Gemini for Science package, which unites several of the company’s LLM-based scientific systems under one brand.

This includes the hypothesis-generating AI Co-Scientist and algorithm-optimizing AlphaEvolve, which are still not publicly available—but as Google is now allowing any researcher to apply for access to Gemini for Science, they may soon see wider adoption in the scientific community. Scientists who were involved in early testing are enthusiastic about their potential: Gary Peltz, a Stanford geneticist, compared using the AI Co-Scientist to “consulting the oracle of Delphi” in a Nature Medicine article.

Gemini for Science isn’t incompatible with specialized tools; to the contrary, agentic systems can be designed to call on such tools when they might be useful. And no agentic system can predict the structure that a protein will fold into without AlphaFold’s help (at least not yet). But the company seems to be shifting its public image—and at least some resources and personnel, such as Jumper—away from specifically developing those kinds of tools. Though it has only been five years since AlphaFold solved the protein-folding problem, both the technology and the discourse have quickly moved beyond that once-revolutionary achievement.

Google has been careful to position this new set of scientific agents as an accelerant for human scientists, rather than a replacement for them—the choice of the name AI Co-Scientist as opposed to AI Scientist, for instance, appears quite deliberate. Hassabis uses that same human-centric framing when he talks about changes in the landscape of scientific AI. “For the next decade or so, we should think about AI as this amazing tool to help scientists,” Hassabis said in an interview published in the Daedalus issue. “Beyond that timeframe, it is hard to say with any certainty, but perhaps these systems will become more like collaborators.”

But no one can be an effective scientific collaborator without also being a skilled scientist in their own right. And if Hassabis is anywhere near the mark when he talks about the “foothills of the singularity,” then AI scientists could eventually exceed the capabilities of their human counterparts.

In a discussion with the journalist Mike Allen at I/O, Hassabis spoke of how he was initially inspired to pursue AI when he observed how progress in physics had stagnated since the 1970s; he wondered whether the human mind had reached its limits in that domain, and if AI could help to overcome that barrier. Superhuman agentic scientists would certainly fit that bill. We might not ever get anywhere near there, but Google seems to be aiming itself toward that summit.

The Enhanced Games fit right in with the rest of 2026’s longevity vibes

This Sunday, a group of 42 athletes will gather in Las Vegas to compete in a somewhat unusual sporting competition. Participants in the inaugural Enhanced Games are being encouraged to take performance-enhancing drugs. The goal is to “push the boundaries of human performance.”

The games’ organizers have said that competitors will only be taking substances that have been approved by the US Food and Drug Administration, and that they are all being medically monitored and supervised. But they have also said they expect to see world records broken—and are offering substantial prizes to athletes who succeed in doing so.

As you might expect, the event is generating a mix of curiosity, excitement, and condemnation from various quarters. To me, it feels like very much a reflection of where we are today—an era of peptide-crazed looksmaxxing in which consumers are being encouraged to get thinner than ever, optimize for longevity, and have their “best baby.” It’s 2026, and if you’re not enhancing, what are you even doing?

So, these games. They’ll feature competitions in four categories: swimming, track and field, weightlifting, and strongman (which also involves lifting weights). Many of the competitors already hold national and world records, and some are Olympic medalists. They’ve been paid a salary and will compete for prizes from a $25 million pot. The money has been a major draw for at least some of the athletes.

Another draw is the opportunity to openly experiment with drugs that might boost their performance. In the world of elite sport, every microsecond and every millimeter counts. Athletes—most of whom arguably have genetics on their side already—follow meticulous diet, training, and recovery protocols and wear specially designed gear that allows them to reach for those performance bests.

But within most sporting communities, there are limits. The World Anti-Doping Agency—an international outfit that fights the use of drugs in sports—maintains a lengthy list of “non-approved substances” that are banned in international sporting events. It features many anabolic steroids (which can build muscle), hormones (such as those that stimulate testosterone production or increase the ability of blood to carry oxygen), growth factors (which can stimulate muscle growth and repair, among other things), and more.

Some of these substances have been FDA approved to treat health disorders. And that means they can be used by participants in the Enhanced Games, according to the organization’s rules.

I’ll briefly point out the obvious here—just because a drug has been approved by the FDA doesn’t mean it’s totally safe for everyone and anyone. The risks associated with use of anabolic steroids, for example, include high blood pressure, acne, depression, and liver tumors. Growth hormone use can cause weak muscles, affect vision, and even lead to diabetes.

“Technological doping,” or using improved equipment to gain advantage, has also been supported by the games’ organizers. Last year, participating swimmer Kristian Gkolomeev was reported to have broken a record in a 50-meter freestyle time trial while wearing a polyurethane “super” swimsuit. Such suits have been banned for use in the Olympics since a slew of record-breaking performances in 2008 and 2009. Back then, the swimming governing body ruled that they gave athletes an unfair advantage. But hey, this is the Enhanced Games, where the word “unfair” seems to have a completely different meaning.

Can we expect more records to be broken on Sunday? Maybe. In addition to prize money for winning an event, any athlete who manages to beat a record stands to win up to $1 million, the sum also awarded to Gkolomeev last year following his time trial. But those performances won’t be recognized by official sporting bodies.

Plenty of concerns have been raised about these games. Some argue that they are unsafe and promote risky drug use. Others see them as a “clown show,” and a slap in the face to “clean” athletes who train hard without the use of prohibited drugs. World Athletics president Sebastian Coe has said that anyone who takes part is “moronic,” and World Aquatics, which oversees international competitions in water sports, has banned Enhanced Games participants from its events and activities.

But. The games—and the participating athletes—will still get a huge amount of attention. As a result, so will performance-enhancing drugs. Enhanced, the company behind the games, also runs an online store. There, you can buy a $52 T-shirt emblazoned with the message “I am Enhanced.”

There is also a range of prescription drugs on offer, including peptides “to support recovery, vitality, and longevity.” One of these is a growth hormone that the FDA approved in 1997 for the treatment of children with “growth failure.” The compounded version offered on the Enhanced website, which is not FDA approved, is marketed for longevity, supporting deep sleep and “overall wellness and vitality.” (“Marketed” is the key word here. The drug has, again, not been approved for that purpose.)

It all fits very well with the zeitgeist. Sure, we don’t yet have any drugs that are designed to extend human lifespan. But the search for anti-aging drugs is getting more attention—and funding—than ever. People, particularly women, are seemingly not allowed to visibly age anymore—we have filters and facelifts for that now. The idea that “death is wrong” is gaining acceptance.

And self-experimentation is rife. “Biohacking” was shortlisted for Collins Dictionary’s Word of the Year in 2025. Peptides are everywhere, despite all the unknowns surrounding their safety and effectiveness. So are longevity clinics, despite the fact that most are selling unproven treatments. US states like Montana are making it easier for people to get hold of unapproved “therapies.”

Companies are even offering would-be parents the option to choose the potential future children expected to live longest. Yep—you can supposedly optimize your embryos now, too.

In this climate, the Enhanced Games don’t feel so radical. They feel entirely fitting for our era of questionable optimization despite the risks —an era when, apparently, being human is no longer enough.

Structural connectome analysis using a graph-based deep model for prediction of non-imaging variables

We address the prediction of non-imaging variables based on structural brain connectivity derived from diffusion magnetic resonance images, using graph-based machine learning. We predict age and the mini-mental state examination (MMSE) score as examples of a demographic and a clinical variable. We propose a machine-learning model inspired by graph convolutional networks (GCNs), which takes a brain connectivity graph as input and processes the data separately through a parallel GCN mechanism with multiple branches. The novelty of our work lies in the model architecture, especially the Connectivity Attention Block, which learns an embedding representation of brain graphs while providing graph-level attention. We show experiments on publicly available datasets of PREVENT-AD and OASIS3. We validate our model by comparing it to existing methods and via ablations. This quantifies the degree to which the connectome varies depending on the task, which is important for improving our understanding of health and disease across the population. The proposed model generally demonstrates higher performance especially for age prediction compared to the existing machine-learning algorithms we tested, including classical methods and (graph and non-graph) deep learning.

Associations between serum tumor necrosis factor-alpha, hippocampal-prefrontal-ventricular volumes, and clinical profiles in schizophrenia: a biomarker and neuroimaging study in North Sumatera, Indonesia

IntroductionSchizophrenia is a neurodevelopmental disorder with systemic inflammation that is thought to play a role in its pathophysiology. The pro-inflammatory cytokine tumor necrosis factor-alpha (TNF-α) has been reported to be dysregulated in individuals with schizophrenia, but its relationship with brain volume changes remains inconsistent. This study aimed to analyze differences in TNF-α levels and brain volume (hippocampus, prefrontal cortex, and lateral ventricles) and their correlations in schizophrenia patients compared to healthy controls.MethodsThis analytical cross-sectional investigation was conducted on 50 schizophrenia patients and 50 healthy Batak controls at the Prof. M. Ildrem Provincial Mental Hospital, Medan. Serum TNF-α levels were measured by ELISA, while brain structure volume was measured by 1.5 T MRI and analyzed using AnalyzePro software.ResultsTNF-α levels in the schizophrenia group were significantly lower (3.35 pg/dl) than in the control group (16.90 pg/dl). No significant differences in hippocampal and prefrontal cortex volume were found between the groups. However, lateral ventricle volume was significantly larger in schizophrenia. Correlation analysis showed a weak negative relationship between TNF-α and prefrontal cortex volume only in the schizophrenia group.DiscussionThe finding of low TNF-α in schizophrenia supports the complexity of immune dysfunction in schizophrenia, which may be influenced by medication or disease phase. Consistent ventricular enlargement reinforces previous neuroanatomical findings. The association of TNF-α with the prefrontal cortex only within the schizophrenia cohort indicates a specific interaction between inflammation and brain morphology in this patient population.ConclusionsThe findings support TNF-α dysregulation in schizophrenia, albeit at lower levels. Lateral ventricular enlargement was consistently found, while changes in hippocampal and prefrontal cortex volume may be more subtle in this population. The negative correlation between TNF-α and the prefrontal cortex in schizophrenia suggests a potential role for inflammation in the pathology of this brain region.

Baseline working memory was associated with improvement in psychological quality of life in patients with persistent depressive symptoms: a prospective observational study

BackgroundCognitive dysfunction is prevalent in patients with depressive symptoms and contributes to impaired quality of life (QOL) and functional outcomes. However, the prognostic significance of specific cognitive domains for long-term outcomes remains unclear in patients with persistent depressive symptoms.MethodsIn this prospective observational study, 119 patients completed the detailed examination program in routine clinical care. Of these, 84 patients with persistent depressive symptoms provided consent for study participation and met the eligibility criteria for inclusion in the present study, including completion of the baseline assessment with the Wechsler Adult Intelligence Scale, Fourth Edition (WAIS-IV). The diagnostically heterogeneous cohort included patients with major depressive disorder/dysthymia as well as bipolar and related disorders. The World Health Organization Quality of Life Instrument, Short Version (WHO-QOL-26) and the World Health Organization Disability Assessment Schedule 2.0 were assessed at baseline, three, and six months; 60 participants provided 3-month follow-up data and 50 provided 6-month follow-up data. Primary hierarchical regression analyses were exploratory and conducted using complete-case data.ResultsAt baseline, participants exhibited relatively higher verbal ability and lower processing speed compared to normative data. While no significant group-level improvement in QOL was observed over six months and functioning did not improve overall, the WHODAS 2.0 standardized total score was temporarily higher at 3 months than at baseline. In exploratory analyses of complete cases, higher baseline working memory was significantly associated with greater improvement in the psychological domain of the WHO-QOL-26 at six months (β = 0.40; ΔR² = 0.14; p < 0.01). No other cognitive domains showed such associations.ConclusionsWorking memory was associated with subsequent improvement in psychological well-being and may represent a candidate prognostic marker in patients with persistent depressive symptoms. Given the exploratory nature and modest sample size, these findings require replication in larger, diverse populations.

Tech researchers are suing the Trump administration over the future of online safety

Since its earliest days back in office, the Trump administration has been going after researchers who study and try to counter hate speech, harassment, propaganda, and disinformation online. 

Now, some of those researchers are fighting back. Last week their lawsuit—which could have global repercussions for online safety and free speech—made its first appearance in court. 

This fight started a year ago, when US Secretary of State Marco Rubio announced on X what he called a “visa restriction policy” against “foreign officials and other persons” who were “complicit in censoring Americans.” Since then, a handful of foreign officials and researchers have been barred from travel to the US, and in theory, anyone working in fact-checking or online trust and safety more broadly could face the same restrictions. 

Still, the exact implications of Rubio’s announcement are unclear—purposefully so, argues Carrie DeCell, a lawyer representing the researchers. “This policy is expansive and incredibly vague, and the chilling effects are correspondingly enormous,” DeCell said outside the courthouse in Washington, DC, on May 13.  

The case has been brought by the Coalition for Independent Technology Research (CITR), an advocacy organization for tech researchers. It is suing Rubio, former US secretary of homeland security Kristi Noem, and former US attorney general Pam Bondi and asking the court to strike down the policy as unconstitutional. In their complaint, the plaintiffs say the policy violates the speech and due process rights of foreign-born tech researchers and workers whose “work supports greater moderation of content on the [tech] platforms.”

CITR is represented by Columbia University’s Knight First Amendment Institute and the legal nonprofit Protect Democracy. DeCell, a senior staff attorney at the Knight Institute, tells MIT Technology Review that they’re in court because the Trump administration is effectively “using immigration law to punish people for expressing views that it disagrees with.” 


This story is part of MIT Technology Review’s “America Undone” series, examining how the foundations of US success in science and innovation are currently under threat. You can read the rest here.


Most immediately, the plaintiffs are asking the government to halt these visa restrictions while the case proceeds. Zachariah Lindsey, the assistant US attorney representing Rubio and the other defendants, argued in last week’s hearing that the government is not targeting speech but, rather, “conduct [that] is assisting or facilitating foreign government censorship of free speech.” At the end of the week, the government filed a motion to dismiss the case.

The judge has yet to rule on either motion, and his questions so far appeared to focus on parsing what (and who) is actually affected by the State Department’s announcements, as well as other procedural issues. 

The outcome of the case may ultimately affect how much the public knows about the risks of social media and AI, says Nicole Schneidman, head of Protect Democracy’s technology and data governance team. The workers bringing this suit, she says, “serve a really, really important function in educating the public, holding tech companies accountable, doing research on the ramifications that advanced technology has on our society.” 

“A political witch hunt”

CITR’s lawsuit is the latest salvo in a yearslong battle over how the internet should be moderated, and by whom—a question that has become increasingly political and entangled in allegations of censorship. 

For years, Trump and his allies have claimed to be victims of a vast conspiracy between government agencies, civil society groups, academics, and Big Tech platforms to specifically censor conservative voices online. According to this narrative, a so-called “censorship-industrial complex” helped the Biden administration subvert First Amendment protections on speech by allegedly outsourcing censorship to these groups.

The State Department claims Rubio was able to implement the immigration policy because the Immigration and Nationality Act authorizes him to “render inadmissible any alien whose entry into the United States ‘would have potentially serious adverse foreign policy consequences for the United States.’” Before the current Trump administration, the statute was rarely invoked, and when it was, it was typically with more limited, specific criteria, rather than its current application against anyone who has participated in alleged censorship—an action that has no legal definition. 

The administration first deployed the policy in July 2025, when Rubio issued a statement announcing the revocation of visas for Alexandre de Moraes, the lead justice on the Brazilian Supreme Federal Court, and “his allies on the court” who were involved in prosecuting Jair Bolsonaro, Brazil’s former president. The prosecution was a “political witch hunt,” said Rubio, calling it evidence of a “censorship complex so sweeping that it not only violates basic rights of Brazilians, but also … targets Americans.”

Then, in early December, the State Department issued instructions to embassies to reject H-1B visa applications from individuals who had worked specifically in fact-checking, online trust and safety, and mis- or disinformation research, as Reuters first reported. 

A few weeks later, on December 23, the agency announced visa restrictions for five Europeans whom it accused of censoring Americans. This included two CITR members: Imran Ahmed, founder and CEO of the Center for Countering Digital Hate, which documents hate speech on social media platforms, and Clare Melford, cofounder of the Global Disinformation Index, which ranks websites according to how often they publish hate speech and disinformation. Also banned were the former European Union commissioner Thierry Breton, a key architect of the European Union’s Digital Services Act (which the State Department has called “Orwellian” and an example of censorship), and Josephine Ballon and Anna-Lena von Hodenberg, co-CEOs of HateAid, a German nonprofit that fights online hate speech. 

Ahmed, who lives in the US with his American wife and child, quickly filed his own lawsuit to stave off deportation and halt the policy. A preliminary injunction preventing his detention and deportation is in place as the lawsuit continues. 

The Department of Homeland Security referred questions from MIT Technology Review to the State Department, which referred “specific questions” to the Department of Justice, while also writing that “the Trump Administration believes that aliens who are or were involved or complicit in censoring American citizens must face appropriate consequences. An American visa is a privilege not a right.” The Department of Justice did not respond to a request for comment. 

“A gut punch”

Now, more tech researchers are fighting back. 

CITR represents 500 individual and institutional members in 47 countries; 40 are based in the United States, including around 30 noncitizens. The organization argues that US-based tech researchers are experiencing a widespread chilling effect and are having to change or reframe what they’re studying so that it’s less explicitly (or less obviously) about content moderation or countering disinformation. Alternatively, some are leaving the US altogether, or making plans to do so, in order to safely carry out their work. 

CITR member Eirliani Abdul Rahman, a Singaporean online safety expert and a founding member of Twitter’s Trust and Safety Council, is one of these individuals. Her experience was included, though described anonymously, in CITR’s initial legal complaint. 

Back in December 2022, shortly after Elon Musk purchased Twitter, Abdul Rahman and two other Trust and Safety Council members publicly resigned. They spoke out against “red lines” the new owner had crossed, including his reinstatement of accounts that had previously been banned, and noted the marked increase in hate speech on the platform. 

Musk disbanded the council days later, but first he retweeted a post that tagged Abdul Rahman and the others and said: “You all belong in jail.” This led to a level of online harassment, doxxing, and death threats that she had never before experienced. “I was trained as an economist, and I could just see line graphs form in my head of the stochastic jump in what happened,” Abdul Rahman says, referring to the way the dangerous attention spiked after Musk effectively endorsed the other user’s provocation. 

This experience inspired her to pursue a new area of research: using quantitative methods to study and hopefully stop social media harassment “in real time,” she says. 

“The ones that are most harassed are people [who] have historically been marginalized,” she adds. “Most of us know about this already, like it’s intuitive. But until you quantify it, sometimes it’s just not seen and taken seriously.”   

But then Trump was reelected, making the work feel untenable. The US quickly became “a funding desert” for scientific research, she says, with federal support for any research perceived by conservatives to focus on mis/disinformation getting cut. At the same time, tech companies shifted their positions on content moderation to align with the president’s, meaning that her research would be unlikely to have any practical implications: “There’s simply no guardrails around social media anymore,” she says. 

Fast-forward to December 2025, and the travel bans on the five Europeans felt like “a gut punch to the stomach,” Abdul Rahman says. She and Ahmed had both testified earlier in the year before the UK Parliament on the role social media played in spreading false claims about the supposed Muslim identity of a murderer who had killed three British girls; this online activity contributed to violent anti-immigrant and Islamophobic riots across the country in the summer of 2024. 

The targeting of Ahmed and the other Europeans “was the last straw” for Abdul Rahman. Soon after, she left the US for a six-year fellowship in Germany aimed at supporting “international academic freedom”—coincidentally arriving in the country on the same day CITR filed its lawsuit. 

“My body just calmed down,” Abdul Rahman says of landing in Germany. “I didn’t wake up in the middle of the night … always wondering about the next executive order and how it pertained to my situation.”

Abdul Rahman believes this legal battle has implications that reach beyond CITR members and their families. It “pertains to all immigrants in the US to protect our First Amendment rights,” she says.

Additionally, whether fact-checkers, online trust and safety workers, and tech researchers can continue to do their work has a broader impact on anyone who uses the internet. 

Earlier this year, for example, Ahmed’s Center for Countering Digital Hate published widely cited research that Grok’s image-editing feature had generated an estimated 3 million sexualized images, including 23,000 images of children, in an 11-day period. This led to government investigations, lawsuits, and even temporary bans for Grok’s parent company, xAI, across the United States and world. 

“The threats have really sharpened”

MIT Technology Review has reported extensively on this right-wing war on supposed censorship; one of our stories revealing that State Department leadership requested communications records from a now-shuttered office focused on countering foreign disinformation has been included as an exhibit in the CITR lawsuit. This request sought insight into communications with a slew of individuals some far-right activists allege are involved in the “censorship-industrial complex,” including journalists, the German foreign minister, and numerous researchers studying disinformation and hate speech (including Medford, Ahmed, and their organizations).

DeCell tells us that over the past year and a half, there have been more lawsuits against the Trump administration regarding free speech—because “the threats have really sharpened,” she says.

Last year, the Knight Institute sued Rubio on behalf of of university faculty and students who have been arrested, detained, and deported for their pro-Palestinian speech; this past January, a judge ruled that the administration’s deportation policy was unconstitutional. The risk to free speech rights is “palpable” when the government “decides to target people specifically with the threat of rounding them off the streets, throwing them into a detention center, and then potentially deporting them from this country,” DeCell says. 

Though Abdul Rahman is safely abroad for now, she says she’s watching the CITR lawsuit closely. Ultimately, she says, she believes it will determine whether researchers will be able to continue to do their work, “which is to take social media platforms to account,” she says—“making sure there’s actual accountability and independent oversight is critical to protecting our democracies.” 

Exploring the life stories of young adult men in prison with a history of dual harm

People who engage in both self-harm and violence (‘dual harm’) in prison cause widespread disruption to prison services. Whilst the behavioural profile of such individuals is gaining attention, there is very little research which explores their life histories and how these contextualise their dual harm. This study qualitatively explored how five young men in a medium secure prison in England with a history of dual harm (in the community, prison, or both) made sense of their life experiences and engagement in dual harm behaviours. Participants were interviewed using an in-depth life story interview protocol. A narrative analysis identified three themes: ‘Beginning: Making sense of a traumatic childhood’, ‘Middle: Exploring challenges during late adolescence’ and ‘End: Who I am now, and who I must be’. These themes, grounded in life experiences and associated meaning, offer valuable insights into the underpinnings of dual harm behaviours and direction for interventions addressing dual harm in prisons. The paper concludes with a discussion of key implications and directions for future research and practice in relation to the support and management of people who dual harm in custody.

Carolina Aguilar: Brain-Computer Interfaces that Heal Neural Circuits

Jonathan D. Grinstein, PhD, North American Editor of Inside Precision Medicine, hosts a new series called Behind the Breakthroughs that features the people shaping the future of medicine. With each episode, Jonathan gives listeners access to his guests’ motivational tales and visions for this emerging, game-changing field.

Brain-computer interfaces have captured global attention in recent years, but most public discussion has focused on assistive technologies—systems designed to help patients control computers, prosthetics, or digital devices through thought alone. Carolina Aguilar, co-founder and CEO of INBRAIN Neuroelectronics, believes the larger opportunity lies elsewhere: using advanced neural interfaces not just to decode the brain but to treat disease directly. Built around graphene-based technology, INBRAIN is developing implantable systems capable of reading and writing neural signals with far greater precision than traditional metal-based devices.

Before founding INBRAIN, Aguilar spent 13 years at Medtronic, including a decade leading the company’s global neuromodulation business. During that time, she saw both the extraordinary potential of brain stimulation therapies and the limitations of incremental innovation inside large medical technology companies. Her background in neuroscience research, including early work studying the relationship between pesticides and Parkinson’s disease, shaped a long-term interest in circuit modulation and neurotherapeutics. That experience eventually converged with the emergence of graphene as a promising material for next-generation neural interfaces.

In this conversation, Aguilar discusses why INBRAIN chose to focus on therapeutic BCI applications rather than assistive computing, how graphene may enable higher-resolution neural decoding and stimulation, and why Parkinson’s disease became the company’s first major target. She also outlines INBRAIN’s broader vision for personalized neurotechnology, AI-driven therapies, and future applications ranging from epilepsy and memory restoration to bioelectronic treatments for cardiometabolic disease.

This interview has been edited for length and clarity.

 

IPM: You spent more than a decade at Medtronic before founding INBRAIN. What made you realize that incremental innovation was no longer enough, and why did graphene feel like the right breakthrough technology at the right time?

Aguilar: INBRAIN is a graphene-based brain-computer interface (BCI) company developing the most intelligent and adaptive interface between the neural system and AI to solve health for billions.

I started my career in consumer goods, but then I spent 13 years at Medtronic, where, for ten of those 13 years, I was leading the brain stimulation, or neuromodulation, business globally. I was always extremely impressed by what the company vision was and the number of disease areas and patients that we could help. However, I saw that every year we were innovating, but it was a bit incremental versus breakthrough. It’s normal for big companies to prioritize preserving shareholder value and revenue opportunities.

Investing in breakthroughs is harder; they usually just acquire those breakthroughs and then integrate them. And, of course, they revitalized their innovation. So I thought at that point that I could be one of those breakthrough innovators that could take a little bit more risk but then bring much bigger value into the field.

When the right opportunity presented itself to us, after the European Union put €1 billion into bringing graphene to market and after having gone through different speeds as pieces, we realized that it was the time to build INBRAIN and make that breakthrough a step instead of the incremental innovation of the past.

It’s interesting because I actually studied neuroscience at Virginia Tech, and at Virginia Tech, I had to define my thesis to study for my master’s degree. I picked up a combination of pesticides in the study. The study focused on the effects of pesticides and their combinations on brain chemistry in Parkinson’s disease. Right. So it was already quite oriented toward what we are solving today. And I got a grant award for the best vision on circuit modulation. So it was not about targeting the brain. It was about the circuit that was actually managing some of these dynamics of Parkinson’s disease.

At that early stage, I was already very attracted to the problem we are solving today, even though I did not know I would have a company dedicated to that field and to this innovation. However, it was always driving me in that direction. Medtronic was the company that turned that ambition into a real product and broad platform, which I also helped launch globally. So I guess it was always there. It was the seed that had been growing over time.

 

IPM: Medtronic exposed you to commercial, upstream product development, and engineering challenges. What did you learn that prepared you to build INBRAIN, and how did graphene change your engineering perspective?

Aguilar: It also evolves over time. So we start with a problem, such as pesticides in the development of Parkinson’s. Then I think I stepped into Medtronic as a solution to the problem that existed at that time. I was like, great. Now we have a problem. We have a solution. Let’s fix Parkinson’s disease.

At Medtronic, I was exposed to new experiences even while working in a commercial role, as I became a global director. I was exposed to the upstream. At Medtronic, there is a downstream process from which all employees receive their compensation. Then there’s the upstream piece, which is the product development and the planning of the next platforms. I was not in charge of that, but I was exposed to the process, the thinking, what is coming next, how we can make it happen, and where the constraints are. And that intrigued me a lot.

I discovered that I had a little bit of an engineering part inside of me that I never exploited. When the right opportunity came, I received a call from an investor friend of mine, who sometimes asked me to do some small due diligence. And he said, “Hey, there are some guys who have some graphene technology here that you might understand better than me. Do you want to come to the pitch?” I said, “Perfect.”

Then I saw that some of these constraints that we were dealing with in Medtronic in terms of miniaturization, higher-resolution interfaces, charge injection limits, and many of these engineering problems, partially, we could also solve by going graphene and going a different kind of electronics and a different kind of platform. When these consolidated and we created INBRAIN, I had to put a huge amount of effort into understanding semiconductors, microelectronics, mechanical configurations, and data architectures. So it’s been a huge learning journey for me.

 

IPM: INBRAIN appears to be building a platform strategy across multiple applications and devices. What are the product verticals and why does graphene offer capabilities that traditional neural interfaces cannot?

Aguilar: So it’s effectively a platform with three product verticals. When Morgan Stanley released the BCI industry report, it stated that the market was a $400 billion opportunity. We already knew that with one device, you could not capture all that immense market. So we were already working on a stepwise approach to the biggest opportunity. And this is what we have today on the table in the sense that we have a first product. The first product I can show you is a cortical interface. Starting with about 100 contacts, we can grow to a thousand. But in this case, it was not necessary.

It’s made of graphene. So it’s not about the number of contacts but the quality of what these contacts can decode. What is the precision of anatomy and the resolution of disease-related biomarkers that we can decode at high resolution in a way that that resolution is superior to a standard medical technology?

In some cases, standard platinum and iridium are used; in some cases, iridium oxide. We decided to create this first product for use for less than 30 days once it’s approved. It’s not yet approved, but we are getting closer, and it is the one that we actually took first-in-human. The results of those three reports are being generated using this device to demonstrate both the safety of graphene and its decoding superiority compared to metal technology. And, of course, within that first-in-human study, we also decoded speech at the phoneme level.

It was about creating the first step toward making graphene and consolidating advanced materials in clinical practice. So from there, the second product is actually an implantable platform that uses some of that configuration. So we use a cortical and subcortical interface to actually decode a circuit.

In our case, the microcircuit that is involved in Parkinson’s disease. And that is the second one. And the third product is actually the same platform, where instead of these two cortical and subcortical interfaces, we put in the central nervous system. We are truly connected to a vagus nerve interface that decodes the fibers within the vagus nerve, which go into the different modulations of the different organs of the body.

We go from reading the brain to reading the body. This sequence of product verticals opens up the immense possibility of a $400 billion opportunity. But even more important is the number of patients we could actually touch and improve therapeutically with such a platform.

 

IPM: Many BCI companies focus on assistive applications like computer and prosthetic control. INBRAIN appears to be therapeutic. Why did you choose direct disease treatment, and what makes it harder?

Aguilar: People say “invasive” or “noninvasive.” I call them implantable and non-implantable. And within the implantable, there might be different levels of invasiveness, but we were looking at the problem to solve, and especially coming from the field we come from and having experience for ten years, the use cases we saw are assistive BCI, meaning I transfer thought to action in the computer as solving a very important problem, but actually in a small population because, at the end, paraplegia and ALS. They are very important disease areas, but compared to other areas, there are smaller populations.

There were many people and companies, great companies, already aiming to solve that. So Neuralink and Synchron and Paradromics were always positioned there. And we thought, well, we have a similar approach, but maybe with a material that is much better suited to actually read and write bidirectionally, therapeutically. In assistive BCI, you do a ton of decoding. So, you read it and then transfer it to the computer. On the therapeutic side, where we are doing it, you have to read, write, and do that computing already within the implanted system.

It’s a much more complex architecture and a harder problem to solve. Now the benefit is solving the disease for as long as the system is active. And Parkinson’s was only the beginning. We are currently examining the validation of this program. We’re looking at memory restoration. We are looking at a set of disease areas on the cardiometabolic side that I cannot disclose without an NDA that we are developing with Merck KGaA. Those deals were unexplored or suboptimally explored by some of the low-resolution companies in the field.

The post Carolina Aguilar: Brain-Computer Interfaces that Heal Neural Circuits appeared first on Inside Precision Medicine.

The Download: fully artificial chicken eggs and why Musk lost

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.

Colossal Biosciences is growing chickens in a 3D-printed artificial eggshell

The baby chicks were shifting and starting to pip—or trying to hatch. But not from an egg. Instead, these chickens were growing inside transparent 3D-printed plastic cups at the Dallas headquarters of Colossal Biosciences.

The biotech company yesterday claimed it has developed a “fully artificial egg” as part of its effort to resurrect extinct avian species, including birds like the dodo and the giant moa.

Some scientists think Colossal is overstating the breakthrough. But the technology may represent an early step toward artificial wombs.

Read the full story on the science and controversy behind the artificial eggshell.

—Antonio Regalado

Inside the Musk v. Altman Trial

Elon Musk has lost his landmark lawsuit against OpenAI, which centered on allegations that its cofounders Sam Altman and Greg Brockman misled him about the company’s nonprofit mission. But what really happened in the courtroom, and what does it mean for the AI race? 

AI reporter and attorney Michelle Kim, who covered the trial for MIT Technology Review, joined our editor in chief Mat Honan to unpack it all in an exclusive Roundtables discussion yesterday.

Subscribers can watch the full recording now.

MIT Technology Review Narrated: this scientist rewarmed and studied pieces of his friend’s cryopreserved brain

L. Stephen Coles’s brain sits in a vat at a storage facility in Arizona. It has been held there at a temperature of around −146 degrees °C for over a decade, largely undisturbed. Before he died in 2014, Coles had the brain frozen with an ambitious goal in mind: reanimation. 

His friend, cryobiologist Greg Fahy, believes it could be revived one day. But other experts are less optimistic.  

Still, Fahy’s research could lead to new ways to study the brain. And using cryopreservation for organ transplantation is becoming a viable reality. 

—Jessica Hamzelou

This is our latest story to be turned into an MIT Technology Review Narrated podcast, which we publish each week on Spotify and Apple Podcasts. Just navigate to MIT Technology Review Narrated on either platform, and follow us to get all our new content as it’s released.

Can AI learn to understand the world?

The limitations of LLMs are pushing AI researchers towards new systems that understand the physical environment: world models. The likes of Google DeepMind, Fei-Fei Li’s World Labs, and Meta’s former Chief AI Scientist, AI Yann LeCun, have brought this technology to the forefront of AI. 

To explore where this technology is heading next, MIT Technology Review is hosting an exclusive Roundtables discussion on Thursday, May 21, with editor in chief Mat Honan, senior AI editor Will Douglas Heaven, and AI reporter Grace Huckins. Register here to join the session at 19:30 GMT / 2:30 PM ET / 11:30 AM PT.

World models are also one of MIT Technology Review’s 10 Things That Matter in AI Right Now, our list of what’s really worth your attention in the busy, buzzy world of AI.

The must-reads

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

1 Google is changing its search box for the first time in 25 years
Its AI-powered overhaul centers on an “intelligent search box”. (Wired $)
+ “Information agents” will gather information on a user’s behalf. (TechCrunch)
+ Google, Gemini, and Gmail may one day be a single search box. (The Verge)
+ AI means the end of search as we know it. (MIT Technology Review)

2 Samsung workers plan to strike tomorrow over AI profit sharing
They say that their employer isn’t sharing the rewards of the AI boom. (WSJ $)
+ And want ​15% of the company’s annual operating profit. (CNBC)
+ South Korea may invoke emergency powers to stop the strike. (Reuters $)
 
3 The White House is set to release a new executive order on AI safety
It’s slated to launch this week. (Axios)
+ The order seeks early government access to advanced models. (NYT $)
 
4 The FBI plans to buy nationwide access to license plate readers
It wants “data in near real time” from cameras across the US. (Ars Technica)
+ The tech could let it track drivers nationwide. (Newsweek)
 
5 Google will launch a new line of smart glasses this fall
They’re the company’s first attempt since the Google Glass flop. (BBC)
+ Google Gemini will power the interactions with the user. (Guardian)
+ Meanwhile, Anduril and Meta are making smart glasses for warfare. (MIT Technology Review)
 
6 A new bill in Congress proposes a new annual fee for EVs
It could cost drivers an extra $130 a year. (NYT $)
+ The fee will cover highway maintenance costs. (WSJ $)
 
7 OpenAI co-founder Andrej Karpathy has joined rival lab Anthropic
Karpathy was also previously Tesla’s director of AI. (Fortune)
+ He coined the term “vibe coding.” (MIT Technology Review)

8 The fears over Anthropic’s Mythos AI model look overstated
Cybersecurity experts say the hacking threat is exaggerated. (Reuters $)
 
9 Silicon Valley keeps misreading China’s role in tech
Viewing Chinese firms as enemies could do more to hurt than help the US. (Rest of World)
 
10 A book about AI’s effects on truth contains false quotes created by AI
It’s among a spate of controversies involving AI-generated quotes. (NYT $)
+ Yesterday, a lawyer apologised for including them in a court filing. (Reuters $)
+ A senior journalist was recently suspended for using them. (Guardian)

Quote of the day

“It may be that the judges have now awarded a prize to an instance of AI plagiarism—we don’t yet know, and perhaps we never will know.”

—Sigrid Rausing, publisher of literary magazine Granta, casts doubts on the authenticity of the Commonwealth Short Story Prize winners, Wired reports.

One More Thing

SELMAN DESIGN


Who gets to decide who receives experimental medical treatments?

Max was only a toddler when his parents noticed there was “something different” about the way he moved. He was slower than other kids his age, and he struggled to jump. He couldn’t run. A genetic test confirmed their fears: Max had Duchenne muscular dystrophy. 

Desperate to slow its progression, Max’s parents enrolled him in an experimental gene therapy trial. The FDA had approved the medicine on weak evidence—a move that has become increasingly common. 

We urgently need to question how these decisions are made. Who should have access to experimental therapies? And who should get to decide? 

Read the full story on the intense debate over experimental treatments.

—Jessica Hamzelou

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.)

+ Trace the history of the world’s most famous plumber in this biography of Mario.
+ Watch the Earth’s spin in action as a full Milk Moon slowly disappears behind a volcano.
+ This handy tool for movie buffs lets you filter upcoming releases by territory and save them to a local watchlist.
+ A missing cat was reunited with its owner after five years and 270 km apart—all thanks to an old Facebook post.