Is Your Phone Doing to Your Brain What Cocaine Does?

Wolf and colleagues published a 2025 meta-review in Progress in Neuro-Psychopharmacology and Biological Psychiatry pulling together a decade of neuroimaging studies. The finding: heavy smartphone users and chronic cocaine users show overlapping gray matter reductions in the insula, the anterior cingulate cortex, and the orbitofrontal cortex — the brain's centers for craving, self-control, and reward valuation.

The original paper is available via the journal at doi.org/10.1016/j.pnpbp.2025.111248. The overlap is not metaphorical. The same neighborhood of the brain is physically affected. Drugs produce a larger acute impact. Prolonged phone use hits it longer.

The cocaine comparison is not new. A 1999 study by Chang and colleagues in the Journal of Neuropsychiatry and Clinical Neurosciences documented persistent cerebral metabolite changes in the frontal lobes of abstinent cocaine users. The 2025 Wolf review situates the smartphone alongside that same research lineage.

The popular framing — dopamine is the happy molecule, phones give you dopamine hits — misses the actual mechanism. Stanford neuroendocrinologist Robert Sapolsky demonstrated in primate studies that dopamine fires not during a reward, but in anticipation of it. Crucially, when the reward was made unpredictable — a 50/50 chance rather than a guaranteed treat — dopamine release doubled.

Casinos figured this out decades ago. Slot machines are tuned to pay out just often enough to keep the dopamine anticipation system running. The pull-to-refresh gesture on a social media feed is mechanically identical to a slot machine lever. Psychiatrist and neuroscientist Judson Brewer at Brown University has described this directly: "Your phone delivers intermittent reinforcement — the exact reinforcement schedule that makes slot machines addictive. Your phone is a slot machine that lives in your pocket." His clinical research on habit loops and reward-based learning is available at drjud.com.

There is a second half to this mechanism that most coverage omits. Every artificial dopamine spike causes the brain to compensate by lowering baseline sensitivity. After months of variable-reward scrolling, ordinary rewards — meals, conversations, sunsets — stop registering with the same force. Nothing competes. This is how a healthy brain ends up experiencing normal life as flat or boring.

Justin McLeod is the CEO of Hinge, a platform whose entire business depends on users opening it. At Web Summit in November 2019, he said: "I don't think your brain really knows the difference. It's going to get those same neurochemical hits whether you're taking a shot of alcohol, or a line of cocaine, or pulling out your phone to check that latest push notification." The full session — "Abandon your black mirror: Conquering tech addiction" — is archived at vimeo.com/374896245.

Stanford addiction medicine chief Dr. Anna Lembke, author of Dopamine Nation (Dutton, 2021), has described the smartphone as "the modern-day hypodermic needle delivering digital dopamine 24/7." Her clinical research and writing are available at annalembke.com.

Chamath Palihapitiya was Facebook's Vice President of User Growth. At the Stanford Graduate School of Business View from the Top series in November 2017, he said: "The short-term, dopamine-driven feedback loops that we have created are destroying how society works. No civil discourse, no cooperation, misinformation, mistruth." The video is available at youtube.com/watch?v=PMotykw0SIk, timestamp 22:42.

Sean Parker, founding president of Facebook, spoke at an Axios event in Philadelphia the same month: "It's exactly the kind of thing that a hacker like myself would come up with because you're exploiting a vulnerability in human psychology." He added: "God only knows what it's doing to our children's brains." The clip is at youtube.com/watch?v=R7jar4KgKxs.

A 2023 JAMA Pediatrics study from the University of North Carolina at Chapel Hill (Maza et al.) scanned 169 adolescents annually from age 12 to 15. Students who checked social media more than 15 times a day showed growing hypersensitivity in the amygdala, anterior insula, and prefrontal cortex — regions governing social feedback processing and executive control. Students who did not check compulsively showed normal developmental trajectories. The divergence was statistically significant. The full paper is at jamanetwork.com/journals/jamapediatrics/fullarticle/2799812.

The CDC Youth Risk Behavior Survey covering 2011 to 2021 documented the population-level fallout. The percentage of teen girls reporting persistent feelings of sadness or hopelessness rose from 36% to 57%. Teen girls seriously considering suicide rose from 19% to 30%. The full report is at cdc.gov/healthyyouth/data/yrbs/pdf/YRBS_Data-Summary-Trends_Report2023_508.pdf.

In 2023, U.S. Surgeon General Vivek Murthy issued an advisory stating that adolescents spending more than three hours a day on social media face double the risk of anxiety and depression symptoms. The average U.S. teen now spends 4.8 hours a day on these platforms. The advisory is at hhs.gov/sites/default/files/sg-youth-mental-health-social-media-advisory.pdf.

Psychologist Jonathan Haidt has called the 2010-to-2015 period "The Great Rewiring of Childhood" — the shift from a play-based to a phone-based childhood. Face-to-face time among teens fell from 122 minutes per day in 2003 to 67 minutes per day now, according to Bureau of Labor Statistics American Time Use Survey data.

The research is not unanimous. Amy Orben and Andrew Przybylski at the University of Oxford analyzed datasets covering over 300,000 adolescents and found that digital technology use explains approximately 0.4% of the variation in adolescent well-being — smaller than the association with eating potatoes. The paper is in Nature Human Behaviour (2019): nature.com/articles/s41562-018-0506-1.

A meta-analysis by Jeffrey Hancock at the Stanford Social Media Lab similarly found near-zero average effects on well-being.

The key phrase is average effects. Population-level averages compress the distribution. The harm documented in the Wolf brain-imaging review, the JAMA Pediatrics longitudinal study, and the CDC data concentrates in specific subgroups: heavy users, adolescents, nighttime use, passive scrolling on appearance-based platforms. For those groups, the effects documented in the clinical and imaging literature are not near-zero.

On March 25, 2026, a Los Angeles Superior Court jury found Meta and YouTube liable for negligent design and failure to warn in the case K.G.M. v. Meta Platforms and Alphabet. The jury awarded six million dollars total — three million compensatory, three million punitive — with Meta at 70% responsibility and YouTube at 30%. TikTok and Snap settled before trial. Meta and Google deny wrongdoing and are expected to appeal.

The previous day, a New Mexico jury ordered Meta to pay hundreds of millions in a separate child-safety case involving violations of the state's consumer protection laws. Coverage of both verdicts via Reuters at reuters.com/legal/litigation/what-did-jury-decide-social-media-case-against-meta-google-2026-03-25.

The legal significance: for 25 years, Section 230 of the Communications Decency Act shielded platforms from liability for user-generated content. A design-defect claim is a different theory. The LA jury was not deciding whether Meta and YouTube hosted harmful posts — they were deciding whether the architecture of the platforms themselves (autoplay, infinite scroll, algorithmic recommendation, features that reduce stopping points) constitutes a defective product. The jury was persuaded.

At a Senate Judiciary Committee hearing on January 31, 2024, Senator Josh Hawley confronted Mark Zuckerberg directly. Zuckerberg stood, faced the families of victims in the gallery, and said: "I'm sorry for everything you have all gone through. No one should have to go through the things that your families have suffered." The moment is at youtube.com/live/HUjv2Ky7PcM, timestamp 2:35:44.

A February 2025 study by Castro and colleagues in PNAS Nexus (doi.org/10.1093/pnasnexus/pgaf017) blocked mobile internet access — not the phone itself, just the always-on portal — for 467 volunteers for two weeks. Calls and texts still worked. Only 119 participants (25.5% of signups) kept the block active for at least 10 of the 14 days. These were volunteers who wanted to participate and had software helping them.

For those who completed the intervention: 91% improved on at least one measured outcome across well-being, mental health, and sustained attention. The researchers compared the attention effect to reversing approximately ten years of age-related cognitive decline. The depression and anxiety improvements were comparable in effect size to antidepressant medication and cognitive behavioral therapy. Some improvements appeared within the first week.

The brain is not permanently damaged. It is trained. Trained systems can be retrained. The PNAS study provides the most direct quantified evidence for what recovery looks like and how quickly it begins.

The real target is not the phone. It is the feed — the algorithmically ranked, infinitely refreshing, socially loaded scroll that sits inside the app. Camera, maps, messaging, search: none of those are the slot machine. The feed is. Feed Filter exists to remove that layer without requiring you to leave the platforms entirely. See feedfilter.com.