On January 5th 2016, the Federal Trade Commission announced its $2 million settlement with Lumos Labs, the makers of Lumosity. The complaint named specific overclaims: that Lumosity training would sharpen performance in everyday life, delay age-related cognitive decline, and protect against Alzheimer's, dementia, and cognitive impairment associated with health conditions like stroke, traumatic brain injury, PTSD, ADHD, the side effects of chemotherapy, and Turner syndrome. The FTC's case was that these claims went beyond what the company's own evidence supported.
A decade later, the case is still the clearest line in the sand the brain-apps category has been given by a regulator. This post is the honest accounting: what changed, what didn't, and where the line between honest daily practice and overclaim now runs.
What the FTC actually said (and didn't)
A point that often gets lost. The FTC did not say brain training doesn't work. It said Lumosity's specific advertising claims went beyond the evidence the company had. The settlement order specifically named the categories of claim that were out of bounds: real-world performance, age-related decline, dementia, and clinical conditions.
What the FTC didn't rule on:
- Whether daily practice has value as a habit.
- Whether specific cognitive games train specific narrow skills.
- Whether brain training is harmful (it isn't, in any documented way).
The settlement was a regulator's false-advertising finding, not a research finding. The research-side question — does brain training work? — was being argued in the scientific community on a parallel track, and that track produced its own findings.
The 2014 and 2016 consensus statements
Two scientific documents bracket the Lumosity moment.
In 2014, more than 70 cognitive scientists signed the Stanford-Berlin consensus statement — A Consensus on the Brain Training Industry from the Scientific Community. The statement's central argument: the published evidence at the time did not support broad claims that brain-training games improve general cognitive ability or reduce age-related decline. The statement was specific about what the evidence did and didn't show, and it was a notable moment of the scientific community pushing back on commercial overclaim.
In 2016, the Psychological Science in the Public Interest paper by Daniel Simons and colleagues — Do "Brain-Training" Programs Work? — published a comprehensive review of the brain-training literature. The conclusion was carefully framed: brain-training games produce reliable improvements on the specific tasks trained; they produce smaller and less reliable improvements on closely related tasks; and they produce little evidence of broad transfer to general cognitive ability or to real-world outcomes.
Together with the FTC settlement, these two documents formed the post-2016 reference point. Any product in the category had a choice: honest about what the evidence supports, or in the regulatory and scientific crosshairs.
What changed in the category after 2016
Three things changed, in roughly this order.
Marketing language got more specific. The category-wide retreat from "improve your IQ" and "prevent cognitive decline" claims happened over 2016–2018. Read the App Store copy for the major brain-training apps now versus the 2014 archive — the difference is striking. The 2026 versions are mostly specific about what the games train (working memory, attention, reaction time) and quiet about what those gains transfer to.
Honest products narrowed their scope. A subset of products in the category — the post-2016 honest cohort — pulled back from broad-promise marketing into specific, defensible claims about daily practice as a habit, narrow task improvement, and engagement value. The honest cohort is smaller than the broader category, but it's the part of the category that survived regulatory and scientific scrutiny.
The research-side argument continued. The 2016 consensus did not settle the question; it set the floor. Subsequent meta-analyses — including the PMC review on brain-training efficacy — refined the picture. Specific protocols (e.g., dual n-back, working-memory training in some clinical populations) have continued to attract serious research interest. Broad consumer claims still don't hold up.
What didn't change
Three things didn't change in the decade.
The temptation to overclaim is still there. Brain training is a commercial category, and the commercial pressure to promise more than the evidence supports never went away. Several products in the broader category continued to flirt with claims the FTC named, sometimes with subtler language than Lumosity used.
The science-org skepticism remains. Recent coverage like Science.org's analyses of neuroscientists' views on brain training reflects continued careful skepticism. Researchers are open to specific protocols with specific evidence and skeptical of broad consumer promises.
The transfer question is still unresolved. Whether daily practice in cognitive tasks produces measurable real-world performance gains — beyond the trained task itself — is still actively researched and not settled. The honest answer in 2026 is still "more research needed, and most existing studies suggest near-transfer effects are reliable while far-transfer effects are not."
What the honest 2026 products look like
A post-2016 honest brain-exercise product looks like this:
It is specific about what it does. Daily practice of specific cognitive skills. Working memory reps, attention reps, reading reps, math reps, writing reps. Not "boost your brain" or "make you smarter." The product describes the activity, not a promised cognitive outcome.
It doesn't claim to prevent or treat anything. No Alzheimer's claims. No dementia claims. No ADHD treatment claims. The FTC named these specifically; honest products stay on the right side of the line.
It uses everyday language for the underlying skills. Working memory becomes "remember a sequence." Processing speed becomes "react quickly." The translation isn't dumbing down; it's avoiding the implication that the trained surface and the broad cognitive construct are the same thing.
It frames itself as a practice habit, not a treatment. The right comparison class is daily flossing, not medication. The benefit is the habit and the engagement and the daily reps, not a clinical outcome the company is promising.
It includes the caveats. Honest products link to the scientific consensus and the limitations of the evidence, not just the parts of the literature that flatter the product. That includes citing the Simons et al. 2016 review honestly.
Where Senwitt sits on this map
We try to sit cleanly on the post-2016 honest side of the line.
We describe what we do: a seven-minute daily practice block across six skills (writing, math, code, memory, reading, reasoning). We measure private progress against the user's own past. We don't claim it will make you smarter, prevent decline, treat any condition, or transfer to broader cognitive performance in any clinically meaningful way.
What we do claim: daily practice of skills is a reasonable response to a working day where AI assistants are doing increasing chunks of those skills. The evidence base for that argument is the cognitive-offloading literature and the deliberate-practice literature, not the brain-training literature — and it lives on the cognitive offloading research page.
The distinction matters. Senwitt is in a different category from Lumosity 2014. The category Senwitt is in — daily-thinking practice as a counterweight to AI-mediated work — didn't really exist in 2014. It exists now because the cognitive surface underneath AI-heavy work is a 2026 problem that didn't exist a decade ago.
What this is not
A few hedges, because we always hedge here.
This is not a victory lap for the brain-apps category. The FTC settlement was a deserved correction. The category overclaimed for years; the regulator pushed back; the science community pushed back; the honest cohort survived. Calling this the category cleaning itself up would overstate how clean the category actually is.
This is not a claim that brain training is now proven. The science is still doing what it should: refining specific protocols, debunking broad claims, holding the line on transfer evidence. Anyone selling certainty about brain training in 2026 is selling something different from what the science supports.
This is not advice for clinical populations. People with neurological conditions, cognitive impairment, or specific clinical histories should not take brain-app marketing — or this post — as guidance. The clinical decisions are between them and their clinicians.
