Brain Wiring May Explain Why Older Adults Fall for Phishing Emails

Every day, millions of deceptive emails land in inboxes around the world, and a growing body of evidence suggests that older adults are disproportionately likely to take the bait. A new study published in GeroScience offers the first glimpse into the brain’s wiring that may underlie this vulnerability. Researchers led by Chun-Che Hung and Natalie C. Ebner at the University of Florida, working with colleagues at The University of Texas at Dallas, found that age-related differences in the microstructure of specific white matter tracts—long-range nerve fiber bundles that connect brain regions involved in social cognition—partially explain why susceptibility to email phishing increases with age. The findings move the field beyond broad claims about cognitive decline and toward a mechanistic, tract-level account of why some aging brains are more easily deceived than others.

Fraud and financial exploitation of older adults have become a serious public health concern. The Federal Bureau of Investigation’s Internet Crime Report documents staggering losses from online deception schemes each year, and phishing—fraudulent messages designed to trick recipients into revealing sensitive information or transferring money—remains one of the fastest-evolving threats. Previous work by the same research group and others has shown that older age, apolipoprotein E e4 genotype, lower general cognition, and mild cognitive impairment all compound phishing vulnerability. Yet while behavioral risk factors have been mapped in increasing detail, the neural mechanisms that make an aging brain susceptible to deception have remained largely opaque. The new study was designed to fill precisely that gap.

The research team recruited 51 younger adults, averaging about 21.8 years of age, and 46 cognitively healthy older adults, averaging about 68.9 years. Each participant underwent diffusion-weighted magnetic resonance imaging, a technique that measures how water molecules diffuse through brain tissue. Because water moves more freely along the length of axons than across them, the diffusion signal provides an indirect window into the integrity of white matter pathways. The researchers then used deterministic tractography, a computational method that reconstructs the trajectories of fiber bundles from the diffusion data, to isolate a set of a priori tracts of interest that the authors collectively call “social white matter”: the cingulum, the uncinate fasciculus, the anterior thalamic radiation, and the superior longitudinal fasciculus.

These tracts were not chosen arbitrarily. Prior neuroimaging reviews have linked them to core social cognitive processes, including theory of mind, emotion recognition, and the integration of affective and cognitive information during social decision-making. The uncinate fasciculus connects the temporal lobe, including the amygdala, with the orbitofrontal and anterior temporal cortex; the anterior thalamic radiation carries connections between the mediodorsal thalamus and the prefrontal cortex, relaying higher-order information important for learning and decision-making; the cingulum threads through the limbic system; and the superior longitudinal fasciculus links frontal, parietal, and temporal regions supporting attention and executive control. Damage or degradation in these pathways, the researchers reasoned, could impair the ability to detect when a message is not what it appears to be.

To quantify susceptibility, participants completed the PHishing Internet Task, or PHIT, an ecologically valid behavioral paradigm that presents realistic email messages, some legitimate and some malicious, and measures how readily a person is fooled. Prior studies have validated PHIT scores as a meaningful proxy for real-life phishing vulnerability, and earlier work using the task showed that older adults have greater difficulty discriminating safe emails from malicious ones. In the new study, higher PHIT scores—indicating greater susceptibility—were associated with older age, consistent with the broader literature on scam and fraud risk in aging.

From the reconstructed tracts, the team extracted two complementary diffusion metrics. Fractional anisotropy, or FA, indexes the directional coherence of water diffusion; higher values are generally interpreted as reflecting better-organized, more myelinated fiber architecture. Mean diffusivity, or MD, captures the overall magnitude of water movement; elevated MD is typically associated with tissue degradation, such as demyelination, axonal loss, or increased extracellular space. Both measures change systematically with age across the adult lifespan, and both have previously been linked to cognitive decline and, in separate work, to scam susceptibility in community-dwelling older adults.

The central analysis used regression-based mediation models that controlled for sex, education, general cognitive status, and—critically—global white matter integrity, entered separately as global MD and global FA. This statistical approach asks whether the tract-specific measures carry information about phishing vulnerability beyond what can be explained by a person’s age, demographics, overall cognition, or diffuse age-related changes across the whole brain. The mediation analyses were conducted with bootstrap confidence intervals, which do not assume normality of the indirect effect, and sensitivity analyses using ordinal logistic regression produced results consistent with the primary findings.

Two tract-specific signals emerged. First, higher mean diffusivity in the right anterior thalamic radiation partially mediated the relationship between age group and phishing susceptibility: older participants showed elevated MD in this tract, and elevated MD was associated with higher PHIT scores. Second, lower fractional anisotropy in the left uncinate fasciculus also partially mediated the age-to-susceptibility pathway, with reduced FA signaling weaker microstructural organization in this limbic-prefrontal connection. Because the effects were partial, other factors beyond these two tracts clearly contribute to age-related phishing vulnerability, but the results establish that the integrity of specific social white matter pathways accounts for a meaningful share of the age difference in deception susceptibility. Alternative analyses in which control tracts were entered last rather than first yielded nearly identical results, strengthening confidence in the specificity of the findings.

The two tracts tell complementary stories about how the aging brain may become easier to deceive. The right anterior thalamic radiation is part of cortico-limbo-thalamo-cortical circuitry implicated in executive function and higher-order decision-making, and degradation along this pathway could weaken the analytical control needed to scrutinize an email’s claims. The left uncinate fasciculus, by contrast, links emotional and memory-related temporal regions with prefrontal evaluative systems, and its deterioration could blunt the integration of affective signals—such as the subtle unease that a suspicious message should provoke—with reasoned judgment. Phishing emails are engineered to exploit precisely these channels, deploying urgency, authority, and emotional pressure to short-circuit careful evaluation. A brain whose social white matter has degraded may be less equipped to detect the mismatch between what a message says and what it truly intends.

The authors frame their results as advancing a social white matter-based account of age-related exploitation risk, one that complements existing models emphasizing cognitive, social, and neural determinants of diminished decision-making in aging and dementia. Because the study is cross-sectional, comparing younger and older groups rather than following individuals over time, it cannot prove that white matter decline causes phishing susceptibility; longitudinal work will be needed to establish temporal precedence. The larger project also included middle-aged adults for whom MRI data were not collected due to budget limitations, leaving the middle of the adult lifespan unexamined here. Nevertheless, the study is the first to connect tract-level social white matter integrity to a validated behavioral measure of phishing vulnerability, and all data and analysis code have been deposited in the Open Science Framework for independent verification. If the findings hold up, they point toward a future in which diffusion imaging could help identify individuals at elevated risk of online fraud before the damage is done, and in which interventions targeting the integrity or function of these social pathways—from cognitive training to vascular health measures—might help protect a growing population of older internet users from the predators who target them.

Subject of Research: White matter tract integrity and phishing susceptibility in aging

Article Title: The role of social white matter tracts in susceptibility to phishing in aging

Article References: Hung, C.-C., Lin, T., Lung, T.-C., Faulkner, R., Kennedy, K. M., & Ebner, N. C. (2026). The role of social white matter tracts in susceptibility to phishing in aging. GeroScience. https://doi.org/10.1007/s11357-026-02574-4

Image Credits: AI Generated

DOI: 10.1007/s11357-026-02574-4

Keywords: aging, phishing susceptibility, white matter, diffusion MRI, fractional anisotropy, mean diffusivity, uncinate fasciculus, anterior thalamic radiation, social cognition, fraud vulnerability, GeroScience, deterministic tractography