Biohacking: what's real, what's risky, and what AI is about to change
Oct. 21, 2026 · The Science Bits
"Biohacking" covers everything from going to bed on time to injecting a powder you bought from a website that says "not for human use." Those are not the same hobby. Here's why people do it, where the real danger lives, where it's all heading, and why AI is about to pour gas on the whole thing, for better and for worse.
First, what are we even talking about?
Biohacking just means running experiments on your own body to feel better, perform better, or live longer. The trouble is that one word stretches across a huge range of risk, from "harmless" to "please don't."3 Think of it like a ladder:
- Low riskTracking sleep with a ring or watch, lifting weights, sauna, fasting windows, regular bloodwork, cutting junk food.
- MiddleStacks of supplements and "nootropics," glucose monitors when you're not diabetic, red-light panels, pricey full-body scans, at-home brain-stimulation headsets.
- High riskInjecting unapproved peptides from the gray market, buying "research chemicals," DIY gene therapy, self-implanted devices.
The headlines love the top rung: tech millionaires and their injection cocktails. A lot of real biohackers spend their time on the bottom rung.
Why people do it
This part surprised me. Outliyr, a longevity website, runs an ongoing survey of its own readers. In its latest update, the largest age group was 65 and up, the most common habit was plain old strength training, and only about 1 in 100 said they spend over $1,000 a month. About a third said they "just want to feel good now" rather than chase a long lifespan.1 One big caveat: that's only 75 people, all from a community of committed health nuts, so it's a snapshot of the hobby, not of America. Still, it pokes a hole in the "it's all rich tech bros" picture.
The other reasons that keep coming up:
- Feeling brushed off by doctors. Researchers who interviewed people ordering their own genetic risk reports heard a common thread: they felt their health worries had been dismissed, so they went looking on their own.2
- Wanting control. Your body is the one machine you'd most like to run yourself, and gadgets make it feel like you can.
- Curiosity and tech comfort. If you already tinker with PCs, tinkering with your sleep score scratches the same itch.
- Influencers. Researchers point to social media and wellness personalities as a big reason risky products start to feel normal.3
The ups
What's actually working
- The boring stuff wins. Sleep, exercise, and real food are the "biohacks" with the most evidence behind them, and wearables are genuinely good at nagging you into them.
- Peptides as medicine are legit. There are more than 100 FDA-approved peptide drugs. The science is real; it's the untested, unregulated versions that are the problem.4
- The industry is trying to grow up. Market analysts describe a slow shift away from wild self-experiments toward things that can actually be measured and backed by evidence.21
The dangers
The biggest biohacking story of 2026 is injectable peptides, so it's worth slowing down here.
Peptides are short chains of amino acids. Your body makes its own and clears them out within minutes. The lab-made "wellness" versions are chemically altered to stick around much longer, and that can get your immune system involved.4 The popular ones have very little human research behind them. The U.S. Anti-Doping Agency says flatly that no human clinical trials have shown BPC-157 works for anything, no safe dose is known, and athletes are banned from using it.7
Then there's where people get them. For many users the main source is the gray market: online sellers who label vials "for research use only" to sidestep drug rules.5
What can go wrong with gray-market vials
- The wrong compound, the wrong amount, or leftover junk from manufacturing4
- Bacterial toxins. One independent testing lab ran 140 endotoxin tests on peptide samples and found nearly 8% had a measurable amount, which matters because these get injected.6
- No real oversight of how the product was made, stored, or shipped
- A failed drug test if you compete in sports7
Where the law stands right now
This changes fast, so here's the timeline as of this week:8
- 2023: The FDA put BPC-157 and about a dozen other peptides on a list of ingredients that raise significant safety concerns, which kept compounding pharmacies from making them.
- April 2026: The FDA said it would pull 12 peptides off that list. That removed the safety flag, but it did not make them legal for pharmacies to make.9
- July 2026: An FDA advisory committee voted to recommend six of seven peptides for pharmacy use, including BPC-157. That went against the FDA's own staff scientists, who said the evidence wasn't there.10 The BPC-157 vote was 8 to 6, with one abstention.11
- August 2026: The FDA sent warning letters to five online peptide sellers and made clear a "research use only" label doesn't get them off the hook.
- Oct. 14, 2026: An "interim access" policy for some peptides went to the White House budget office for review. It hasn't been issued.
Bottom line: none of these peptides is an FDA-approved drug, and none has been added to the official pharmacy list yet. People on one side argue a regulated route would pull buyers off the gray market. People on the other side worry it will make unproven products look official.12 Both have a point.
DIY gene therapy
At the very top of the ladder is people editing their own genes. Back in 2017, one biohacker injected himself with a gene-editing tool on stage at a conference, and another livestreamed injecting himself with an untested HIV treatment. The FDA warned that selling do-it-yourself gene therapy kits is against the law. Experts said home-made versions probably weren't strong enough to do much, but could still set off an immune reaction.13
The quiet one: worrying yourself sick
Tracking can backfire. Sleep researchers even have a name for people who get so fixated on a perfect sleep score that their sleep gets worse: "orthosomnia."14 More data isn't always more health.
How AI could change it
This is where it ties back to what this site is about, and honestly it cuts both ways.
The good side
- Making sense of the data. A ring and a watch spit out a mountain of numbers. Google researchers built an AI coach that reads wearable data, and on sleep and fitness exams it scored as well as or better than a sample of human experts.15
- Gadgets that react. The newer wave of wearables is moving from just tracking to actively adjusting things, like stimulation or sleep settings, in real time.22
- Real research, faster. AI tools that predict and design protein shapes won the 2024 Nobel Prize in Chemistry. That's how future peptide drugs that are actually tested will get made.23
- A checker, not a cheerleader. Used right, AI can explain what a study really found and help you write better questions for your doctor.
The bad side
- Confident wrong answers. A BMJ Open study tested five popular chatbots on health topics, including nutrition and athletic performance. About half the answers were problematic, roughly 1 in 5 badly so, and none of the chatbots produced a fully accurate reference list.16
- Missing emergencies. An editorial in eClinicalMedicine pointed to an independent safety check that found ChatGPT Health under-rated the urgency of 52% of emergency scenarios.17
- No better than going it alone. In an Oxford study of about 1,300 people, those using chatbots made no better health decisions than people using regular searches or their own judgment.18
- Harm nobody sees. A paper in Nature Health made a sharp point: bad chatbot advice happens in a private chat, so doctors and regulators may never find out.19
- It's already mainstream. A KFF poll found about 1 in 3 U.S. adults used AI for health information in the past year. Most who used it trusted it, and only about 6 in 10 of those asking about physical health followed up with a doctor.20
Put those together and you can see the risk: someone asks a chatbot how to "stack" peptides, gets a confident answer with a made-up reference attached, and orders a vial that same night. AI didn't build the gray market, but it can make it feel a lot more official.
Where it's heading
Into the home. Expect more rings, watches, scales, and glucose sensors, all feeding one AI "coach" app.
Big money, fuzzy numbers. One market report sizes biohacking at about $32.5 billion this year; another puts it closer to $2 billion.21 When estimates are that far apart, nobody really knows, but everyone agrees it's growing.
The peptide showdown. The FDA's next move decides whether these peptides land in pharmacies with a doctor involved or stay in the gray market. Five more peptides are expected to get their own committee review by early 2027.8
AI on the hook. Researchers are pushing for AI health advice to be labeled, monitored, and possibly held to malpractice-style accountability.19
What we can learn from it
- Start at the bottom of the ladder. Sleep, movement, food, and bloodwork give you the most for the least risk.
- Data is a clue, not a diagnosis. A bad sleep score is a reason to pay attention, not to panic.
- "For research use only" means "not checked for you." If a seller won't stand behind putting it in a human, don't put it in yours.
- Use AI as a research buddy, not a doctor. Ask for sources, then actually open them. Bring what you find to a real doctor.
- The biohackers have a point. A lot of them got into this because they felt ignored. Being curious and taking charge of your health is a good thing. Just do it with your eyes open.
My take: I get the appeal. I'm a tinkerer, and the idea of tuning your body like you'd tune a PC is hard to resist. But a PC you can reimage when you mess it up. Your body you can't. The honest truth is the stuff that works best is the least exciting, and the stuff getting the hype has the least proof behind it. AI is going to make biohacking way more popular, because it makes all that data feel personal and smart. That's great when it's nudging you to bed an hour earlier. It's dangerous when it's confidently telling you what to inject. Use it to ask better questions, not to skip the doctor.
How this was made: I used AI (Claude, by Anthropic) to help research and draft this piece. Every fact was then checked against the original source listed below, and everything is written in our own words. Numbers and quotes belong to the people and outlets credited. This is a look at the science and the news, not medical advice. Talk to a doctor before you put anything new in your body.
References & credits
- Outliyr, "State of Biohacking & Longevity" (ongoing reader survey, updated May 2026).
- Lowes, K., Borle, K., Folkersen, L., & Austin, J., "A Qualitative Study Exploring the Consumer Experience of Receiving Self-Initiated Polygenic Risk Scores from a Third-Party Website," medRxiv (2022).
- Observer Research Foundation, "Biohacking, Peptides, and the Grey Market" (Apr. 24, 2026).
- Skeptic, "The Peptide Craze: Biohacking and Human Guinea Pigs" (2026).
- Personal Care Insights, "US FDA panel weighs injectable peptides as experts demand gray market guardrails" (July 2026).
- Finnrick, "Why endotoxin testing matters for peptides" (Feb. 6, 2026).
- U.S. Anti-Doping Agency, "BPC-157: What Athletes Should Know."
- Holland & Knight, "Peptides at a Crossroads: What Wellness Providers Need to Know About HHS and FDA Actions" (Oct. 16, 2026).
- Frier Levitt, "FDA to Remove 12 Popular Peptides from the Category 2 'Do Not Compound' List" (Apr. 16, 2026).
- TIME, "An FDA Committee Just Voted in Favor of Peptides, Despite the Agency's Opposition" (July 23, 2026).
- Regulatory Affairs Professionals Society, "FDA advisory committee backs two controversial peptides" (July 2026).
- Reason, "The Peptide Gray Market Is Already Here. The FDA Is Just Catching Up." (July 27, 2026).
- MIT Technology Review, "Biohackers Disregard FDA Warning on DIY Gene Therapy" (Dec. 1, 2017).
- Sleep Foundation, "Orthosomnia," citing Baron, K. G., et al., "Orthosomnia: Are Some Patients Taking the Quantified Self Too Far?" Journal of Clinical Sleep Medicine (2017).
- Google Research, "A personal health large language model for sleep and fitness coaching," Nature Medicine (2025).
- Epocrates, "Confident answers, shaky facts: popular AI chatbots put to the test," summarizing a BMJ Open study (Apr. 2026).
- eClinicalMedicine editorial, "The risks of AI-generated health advice" (Mar. 2026), via HIFA.
- University of Oxford, "New study warns of risks in AI chatbots giving medical advice" (Feb. 10, 2026), on a study in Nature Medicine.
- Epocrates, "Hidden harms: When chatbot health advice goes wrong," summarizing a Nature Health perspective (Oct. 2026).
- KFF, "KFF Tracking Poll on Health Information and Trust: Use of AI for Health Information and Advice" (Mar. 25, 2026).
- Global Market Insights, "Biohacking Market Size" (May 2026); Research and Markets, "Biohacking Market: Global Forecast 2026–2032" (Sept. 2026).
- Outliyr / High Performance Longevity podcast, "2026 Biohacking & Longevity Trends" (2026).
- The Royal Swedish Academy of Sciences, "The Nobel Prize in Chemistry 2024" (Oct. 9, 2024).
Share image: original artwork made for this article. No outside images used.
More to read
- From MKUltra to AI: the long history of messing with your mind
- Forced to pick, two big AIs split along party lines
- Real race, fake videos: AI ads take over the Texas Senate fight
- The first AI election: $137 million in AI-made ads, and the AI money behind the races
- How to spot an AI political ad before you vote (or share it)
- What Trump is actually doing about AI, and what Americans think of it