Brain training is not a scam, but the way it has often been advertised was. The category has real, replicable, well-documented effects on the specific cognitive tasks it trains. It does not have the broad cognitive benefits the worst marketing has claimed. The honest current state of the science is narrower than the press cycle suggested, and wider than skeptics often allow. This piece walks through what the evidence actually shows, what got the industry into trouble with the FTC, and how to tell honest claims from marketing.
"Brain-training interventions improve performance on the trained tasks."
Simons et al., 2016, Psychological Science in the Public Interest
Where the skepticism comes from
The current public skepticism about brain training is not unearned. It has three real sources.
1. The FTC settlement. In January 2016, the Federal Trade Commission settled with Lumos Labs, the maker of Lumosity, for $2 million. The complaint specifically cited claims that the games could "stave off mild cognitive impairment, dementia, and Alzheimer's disease," and could "help users perform better at work and school." The FTC found the company had not adequately substantiated those claims. The original suspended judgment was $50 million; only the financial state of the company kept it from being collected.
2. The 2014 Stanford consensus letter. That same year, more than 70 cognitive psychologists and neuroscientists, organized by the Stanford Center on Longevity and the Max Planck Institute for Human Development, signed a public statement saying the marketed benefits of brain-training products were "exaggerated and at times misleading." The letter became the field's reference point.
3. The far-transfer literature. The most rigorous published meta-analyses of working memory training, especially Melby-Lervåg, Redick and Hulme's 2016 review covering 87 studies, have not found reliable transfer from working memory training to general intelligence, reading comprehension, or daily-life function.
Brain training reliably improves the trained task. The case for broad cognitive enhancement is much weaker than the marketing implied.
These are real findings, and any honest treatment of the category has to engage with them.
Where the underlying science actually stands
A more careful read of the literature gives a more interesting picture than "scam" or "miracle."
Near transfer is real. If you practice the n-back task, you get better at the n-back task. If you practice digit-span, you get better at digit-span. This is uncontroversial and replicates reliably.
Far transfer is contested but not zero. The strongest single piece of evidence in the field is the ACTIVE trial (Advanced Cognitive Training for Independent and Vital Elderly). Its 10-year follow-up by Edwards et al. (2017) found that older adults randomized to ten one-hour sessions of speed-of-processing training had a 29% lower hazard of dementia at the decade mark compared with controls. This is not a small effect, and it has not been retracted. It is also one trial, in one specific population, with one specific exercise type.
The FINGER multimodal trial (Ngandu et al., 2015) combined cognitive training with diet, exercise, and vascular monitoring in 1,260 Finnish older adults and found a 25% improvement in overall cognition over two years. The cognitive-training piece was bundled with the rest of the protocol, but the trial set the template for the multimodal approach now under study in the broader brain-health literature.
Most consumer apps probably do less than this. The 2010 Owen et al. Nature study (n=11,430) found that six weeks of generic puzzle training produced near transfer but no far transfer. The 2016 Simons et al. consensus review concluded that brain-training programs "improve performance on the trained tasks, less evidence that such interventions improve closely related tasks, and little evidence that training enhances performance on distantly related tasks or that training improves everyday cognitive performance."
The honest framing: the category is real, the specific protocols matter, and most consumer products are weaker than the strongest research signals.
What the FTC case actually established
The FTC complaint is worth reading in full. It does not say brain training is fake. It says Lumos Labs did not have adequate evidence for the specific health claims it was making. The order prohibits the company from claiming the games:
- Improve performance at school, work, or in athletics
- Delay or reduce age-related cognitive decline
- Reduce risk of cognitive impairment, including dementia and Alzheimer's
- Improve cognitive performance in users with conditions like ADHD, traumatic brain injury, PTSD, or chemo-related cognitive impairment
unless the claims are supported by competent and reliable scientific evidence. That is not a ban on saying anything good about brain training. It is a ban on a specific class of overclaim.
The settlement is, in effect, the FDA-equivalent line for the consumer brain-training industry. Any product that crosses it is risking the same outcome.
How to tell honest claims from marketing
A simple five-question rubric works for any brain-training product:
- What specific cognitive ability does it train? "Brain training" is not an answer. "Working memory under interference," "speed of visual search," "task-switching cost" are answers.
- What dose does it require, and over how long? A protocol that names "10 sessions of 60 minutes" is more credible than "use it daily."
- What outcome is it claiming, in what population? "Reduces dementia risk in adults 65+" is a falsifiable claim. "Boost your brain" is not.
- What independent research supports the claim? Independent meaning not funded by the company. Internal research is fine; it should not be the only research.
- What does the product not claim? Honest products acknowledge limits. Marketing-only products do not.
We use this rubric in our 2026 ranking of brain-training apps. It is the same rubric we apply to ourselves.
What "the right kind" of training looks like
The kinds of cognitive training with the strongest current evidence share three features:
- They train transferable skills. Working memory, processing speed, and attentional control are the three skill domains with the most evidence for near-transfer that scales to broader function.
- They use brief, daily, attention-heavy sessions. Five to fifteen minutes per day, every day. We cover the consolidation neurochemistry behind this separately.
- They progressively adapt to the user's level. Static-difficulty apps plateau within weeks.
Programs with all three features look more like the ACTIVE-trial protocol than like a casual puzzle app. The ACTIVE protocol is what the strongest evidence supports. It is also what most consumer apps are not.
What still doesn't have evidence
Several common claims have weaker support than their popularity implies:
- "Brain age" framing. Apps that estimate a "brain age" younger than your chronological age are doing pseudoscience. There is no validated single number for cognitive function that maps to an age.
- Generic puzzle apps with no progressive adaptation. They make you better at the puzzle. Owen et al. (2010) and Simons et al. (2016) both make this point clearly.
- Memory supplements. Ginkgo, lion's mane in healthy adults, B-vitamin megadosing in non-deficient adults. The clinical-trial evidence ranges from null to mixed.
- Claims that cognitive decline can be undone. No app, supplement, or commercial product has shown this in a peer-reviewed RCT in 2026.
If a product makes any of these claims, you can reasonably treat the marketing as separate from the underlying science.
The basics still outrank the apps
The most honest framing for anyone considering a brain-training product is the hierarchy in our brain-health guide. The interventions with the strongest evidence for cognitive benefit, in roughly this order:
- Aerobic exercise. The most replicated single behavioral intervention in cognitive aging.
- Adequate sleep. Especially deep slow-wave sleep, which is when memory consolidation happens.
- Treating cardiovascular risk factors. Hypertension, high LDL, diabetes. These damage the brain through the vascular system.
- Social engagement and continued cognitively demanding activity. The mechanism is reserve-building.
- Targeted cognitive training. Real, but Tier 3.
If steps 1-4 are missing, no app fills the gap. If they are in place, targeted training plausibly adds incremental, real-world-meaningful gains.
So is it a scam?
The answer is: no, but be skeptical of the claims, not just of the category.
A specific brain-training product is not a scam if it:
- Names the specific skills it trains.
- Cites independent research, including research that disagrees with it.
- Uses a protocol shape consistent with the consolidation literature (brief, daily, adaptive).
- Does not claim to stop dementia, raise IQ, or "rewire your brain."
- Acknowledges that the basics (sleep, exercise, vascular health) outrank it.
A specific brain-training product is closer to a scam if it:
- Promises broad cognitive boosts without naming the protocol or population.
- Uses "brain age" gimmicks.
- Implies transfer to daily-life function without RCT support.
- Avoids mentioning the FTC's 2016 settlement when comparing itself to "what brain training can do."
The category includes both kinds. The skill is telling them apart. For the broader picture, see our complete cognitive training guide.
A practical bottom line
- Brain training is real. The marketing has often not been.
- The strongest evidence is for specific protocols in specific populations.
- The FTC's 2016 settlement is the line for what apps can claim.
- A five-question rubric (skill, dose, outcome, population, independent research) separates honest products from marketing.
- Sleep, exercise, and cardiovascular care outrank any app for cognitive benefit.
- Be skeptical of the claims. Don't be skeptical of the underlying science.
That's the honest version.