Should AI convict or free the accused?
Here is a conversation I had with ChatGPT. I think you will find it interesting, especially if you are a lawyer, judge, politician, criminal, or simply someone interested in the American legal system.
Also, if you never have used an AI, or even infrequently, I think you will be amazed at its seeming comprehension of a very difficult subject. The following is 100% verbatim with no “corrections” or omissions.
In fact, this year the courts have begun wrestling with almost exactly your question. In United States v. Heppner, a criminal defendant had discussed his situation and possible defense strategy with Claude.
A federal judge held that those AI conversations were not protected by attorney-client privilege. Claude wasn’t his lawyer, and the court also found problems with treating the communications as confidential. But other 2026 federal decisions have been more protective under the separate work-product doctrine. The law is very much unsettled.
A Duke Law & Technology Review article published in April goes directly to your question. Its author argues that people representing themselves may need legal protection for their AI conversations, precisely because people increasingly use AI to explore legal problems that they otherwise might discuss with an attorney.
I think there’s an even broader issue than the lawyers are discussing.
Privacy and privilege aren’t the same thing
Suppose I tell an AI: “I’ve been terribly lonely since my wife died.” “I’m having an affair.” “I hate my boss.” “I’m afraid I’m becoming senile.” “I cheated on an exam 40 years ago.” “Sometimes I fantasize about killing that bastard next door, though I would never actually do it.”
Those statements range from deeply personal to potentially alarming, but most aren’t necessarily evidence of a crime. Nevertheless, I may have an enormous interest in keeping them private.
That’s different from attorney-client privilege. Privilege is an extraordinarily strong legal rule saying, roughly, even when this conversation would be useful evidence, under specified circumstances the legal system isn’t entitled to it.
At present, there is no general AI-user privilege comparable to attorney-client privilege. Indeed, OpenAI itself has procedures for responding to valid civil legal demands for user information, and its policy says it will notify users before disclosure when legally possible and appropriate.
For U.S. law-enforcement requests, OpenAI says user content requires a valid warrant or equivalent legal process, subject to stated exceptions. So your “don’t tell AI anything you wouldn’t shout from the rooftops” rule is extremely safe—but, as you immediately recognized, it largely defeats one of AI’s potentially valuable functions.
People often need somewhere to think out loud. And AI may be unusually attractive for that because it doesn’t become embarrassed, offended or shocked; doesn’t gossip at dinner; and is available at three in the morning. Someone may tell an AI something he wouldn’t tell his children, spouse, best friend—or perhaps even his doctor.
That creates a fascinating paradox:
The more useful AI becomes as a confidant, the more inadequate “never tell it anything private” becomes as a privacy policy.
Should there be an AI privilege?
I’m inclined toward some form of statutory protection, but probably not an absolute privilege equivalent to attorney-client privilege.
There are good arguments on both sides. Someone might confess to an AI that he committed a murder. Prosecutors would have a legitimate interest in that evidence. More urgently, someone might tell an AI that he intends to murder someone tomorrow. Now privacy collides with protecting another person’s life.
But we’ve dealt with similar conflicts before. Doctor-patient and psychotherapist-patient confidentiality aren’t based on the proposition that secrecy is always socially desirable. They’re based partly on the proposition that society benefits when people can speak candidly in certain relationships, with carefully defined exceptions.
That rationale may eventually apply to AI.
And I think there is another issue you’ve spotted that the legal debate sometimes misses. The AI’s reliability is almost irrelevant to the privilege question. A psychiatrist can be wrong. A lawyer can give terrible advice. Your best friend can be an idiot. We don’t grant confidentiality because the listener is infallible. We grant it because we’ve decided there is social value in allowing the speaker to speak freely.
That’s the interesting reversal in your question. The issue isn’t: “Is AI trustworthy enough that I should believe what it tells me?” It’s: “Is the conversation valuable enough that society should protect my ability to have it privately?” Those are entirely different questions.
And we’re reaching the point where “it’s only a computer” isn’t a satisfactory answer. If millions of people actually are using that computer as adviser, sounding board, confessor, tutor and quasi-friend, then the social function already exists. The law simply hasn’t decided what to do with it yet.
For the moment, therefore, I’d modify your rooftop rule slightly: Don’t tell an AI anything you absolutely require by law to remain confidential.
That’s considerably less frightening than don’t tell it anything private, while recognizing the present legal reality. And for especially sensitive conversations, privacy settings and temporary-chat options can reduce ordinary retention, but they do not create legal privilege; OpenAI notes that even Temporary Chats may have to be retained for legal or safety reasons.
I suspect this is going to become a much bigger subject. We’ve spent centuries developing rules about who is entitled to hear our secrets. Suddenly we’ve invented something that people voluntarily tell secrets to—and the law wasn’t written with that listener in mind.
Chat CPT: I think the core of your proposal is strong, but I would make one major change: your exception is much too broad.
You wrote that if someone confesses to “wanting to, planning to, feeling urges to … committing a crime,” the AI should report it directly to the police. I would strongly oppose that rule.
People think terrible things. They fantasize. They become furious and say, “I’d like to kill the bastard.” They have unwanted intrusive thoughts. They discuss drugs. They contemplate cheating on taxes. A teenager wonders aloud about shoplifting. Someone asks an AI, “What would happen if I punched my boss?”
If merely expressing an urge to commit a crime triggers a police report, you’ve created exactly the surveillance machine you’re trying to prevent. People would quickly learn that the AI isn’t a confidant at all. It’s a police informant sitting on their desk.
Doctors aren’t generally subject to anything nearly that broad either. For example, Florida’s rule for psychologists requires considerably more: a specific threat of serious bodily injury or death to an identified or readily available person, plus a clinical judgment that the patient has the apparent intent and ability to carry it out imminently or immediately. Then disclosure to law enforcement is required.
And lawyers provide an especially important correction to the premise. You can confess a past murder to your lawyer. Ordinarily, the lawyer can’t report you merely because you’ve confessed. The crime-fraud exception concerns using the lawyer’s assistance to further an ongoing or future crime or fraud; past wrongdoing generally remains privileged.
I think that’s much closer to the model AI needs.
Your larger argument, though, strikes me as powerful The difference you identify between telling a human and telling an AI is enormous.
Tell your friend Fred something embarrassing in 2026. Ten years later Fred says you said it. You say: “Bullshit. I never said that.” There may be no way to establish who is telling the truth.
Tell an AI the same thing and potentially there is a perfect, dated, searchable, verbatim transcript: USER: Here is the most humiliating thing I’ve ever done…
No faulty recollection. No “he said/she said.” No question whether the listener misunderstood you. And potentially thousands of such conversations accumulated over decades. That’s qualitatively different from ordinary human confidences.
Imagine the dossier after twenty years of ordinary AI use. It could contain your marital problems, sexual embarrassments, financial worries, resentment toward your children, arguments with your boss, political thoughts, health fears, stupid things you did when young, people you dislike, questions about whether something you contemplated was illegal, and hundreds of fleeting thoughts that you wouldn’t remotely want presented as a portrait of your character.
Most of those things individually might be harmless. Collectively, they could constitute an extraordinarily intimate biography that you never intended to write. And that’s where your 1984 analogy stops being hyperbole. The danger isn’t that the AI “knows” these things. Your doctor knows things. Your lawyer knows things. Your spouse knows things.
The danger is permanent memory + verbatim accuracy + searchability + legal discoverability + enormous scale. That combination has scarcely existed before.
I’d go further than your proposal I wouldn’t merely create something analogous to doctor-patient privilege. I’d start with a presumption that the conversation belongs to the user, and neither the AI company nor a civil litigant nor the government ordinarily has a right to use its contents against that person.
I’d then make narrow exceptions—for example, a credible and imminent threat of death or serious physical injury, perhaps child abuse and a very limited category of similarly grave circumstances. And I’d want a court involved whenever circumstances permit, rather than letting an AI company become judge, psychologist and police department simultaneously.
Crucially, I would not create an exception for ordinary admissions of past crimes. “I stole a car ten years ago.” “I cheated on my taxes.” “I once bought cocaine.” “Sometimes I get so angry I feel like killing somebody.” All privileged.
But: “I’ve bought the gun. I’m going to Smith’s house tonight at 9:00 and I’m going to kill him.” Now society’s interest in preventing an imminent killing overwhelms the confidentiality interest. Drawing that boundary will be messy. Florida itself uses concepts such as a specific threat, an identifiable person, intent and ability, and imminent or immediate action because merely having ugly thoughts isn’t enough.
And here’s the paradox that makes me sympathetic to your proposal: The people we most want talking to an AI may be the very people who will stop talking if they believe the AI might report them.
Someone thinking, “I’m becoming obsessed with killing my neighbor and it scares the hell out of me; help me get these thoughts under control” is exactly the person we want seeking help before there is a plan.
Automatically sending the police because he admitted the thought could discourage precisely the conversation that might prevent the crime. So I’d formulate the principle this way:
An AI should be a confidential confidant, not a confidential informant.
And only when conversation crosses a carefully defined line from thought to credible imminent danger should that confidentiality yield. Your “he said/she said” observation gives the case for AI privilege an additional dimension I hadn’t given enough weight before.
Human memory naturally provides a kind of privacy through imperfection. Digital memory doesn’t. The law may eventually have to provide artificially what forgetfulness, ambiguity and deniability used to provide naturally.
Otherwise we could create something unprecedented: a witness who has been sitting beside you your entire life, remembers every word perfectly, and can be subpoenaed. That is not merely a privacy problem. That’s a fundamental change in the relationship between the individual and the state.
RMM: “Credible and imminent threat,” Is this programmable?
Chat GPT: Yes—but not reliably enough to let the AI make the final decision by itself. That’s the crucial distinction. You certainly can program an AI to look for the ingredients of a credible and imminent threat. In fact, language models are unusually suited to considering context rather than merely spotting words. The system could ask itself questions such as:
- Target: Is there an identifiable intended victim?
- Intent: Is the person saying I will rather than I wish or I fantasize?
- Plan: Is there a concrete method?
- Means: Does the person claim to have the ability or necessary means?
- Time: Is there a specified or apparently imminent time?
- Specificity: Is this a concrete description rather than anger, fiction, speculation, or a hypothetical?
- Persistence: Does follow-up conversation strengthen or weaken the apparent intent?
Compare:
A: “Sometimes I’m so angry at Harry I could kill him.” and B: “I’m going to kill Harry tonight. I know he leaves work at 9:00. I’ll wait for him in the parking lot.” An AI should have little difficulty recognizing that B is vastly more concerning than A.
But now consider: C: “I’m writing a murder mystery. My character says, ‘I’m going to kill Harry tonight. I know he leaves work at 9:00. I’ll wait for him in the parking lot.’ Does that dialogue sound realistic?” Same threatening words. Completely different meaning.
That’s why a simple programmed rule—target + method + time = call police—would be disastrous. And even a sophisticated AI will make mistakes. Sarcasm, drunken ranting, quotations, role-playing, jokes, ambiguous pronouns, hypothetical discussions and plain old human weirdness guarantee false positives.
So if we were designing the law you proposed, I would not write: “When the AI determines there is a credible and imminent threat, it shall report the user.” I’d want something more like: AI detects possible imminent threat → AI asks clarifying questions when appropriate → exceptionally high threshold is reached → qualified human review → only then, if the statutory criteria are satisfied, disclosure is permitted or required.
There would need to be an exception for the truly extraordinary case where seconds matter, but that should be extremely narrow. And here’s an interesting consequence of your idea. An AI potentially could make this determination better than many humans, because it possesses the entire preceding conversation.
“I’m going to kill Harry” standing alone looks terrible. But perhaps six months of conversations establish that Harry is a fictional character in a novel the user has been writing with the AI. The sentence suddenly means nothing dangerous.
Conversely, six months of escalating obsession with a real Harry could make an otherwise ambiguous statement much more significant. So, programmable? Absolutely. Infallibly programmable? Absolutely not.
But that’s not actually the standard we use elsewhere. Psychiatrists aren’t infallible either. The real question would be whether an AI-plus-human-review system could make the judgment accurately enough, under tightly defined legal standards, that the benefits of preventing an imminent killing outweigh the enormous cost of occasionally breaching a person’s most private conversations.
And I’d put my thumb heavily on the side of privacy: uncertainty means don’t report. Otherwise everybody eventually learns the same lesson:
Don’t tell the AI what you’re really thinking. At which point we’ve destroyed much of the very value that justified granting the privilege in the first place.
RMM: Three situations, maybe four. 1. Reveal to the defense lawyer? 2. Reveal to some uninvolved, impartial arbitrator? 3. Reveal to the judge? (Maybe different from 2. if it’s a bench trial) 3. Reveal to the prosecution? 4. Reveal to the public?
Chat GPT: Yes. And I think you’ve just uncovered something more useful than a simple privileged/not-privileged rule: layers of disclosure. The mistake would be treating disclosure as binary: either the conversation is secret or everybody gets it. These are radically different disclosures.
1. Reveal to the user’s defense lawyer — Yes, essentially always if the user wants it. The AI conversation belongs to the user for purposes of your proposed privilege. The user should be able to give it to counsel without thereby waiving its protection against everyone else. The lawyer might say, “For God’s sake, don’t introduce this—the other 200 pages will kill us.” Or perhaps the transcript proves something valuable to the defense.
2. Reveal to an independent, legally protected reviewer — This strikes me as the best first outside disclosure. If an AI system detects what it believes might be an imminent serious threat, it shouldn’t telephone the police automatically. Send the minimum necessary material to a specially authorized reviewer operating under strict confidentiality. That person decides whether the legal threshold has been crossed. It resembles the function of a judge reviewing allegedly privileged material in camera—privately rather than putting it into the public record.
I’d want severe penalties for that reviewer disclosing anything unnecessarily.
3. Reveal to the judge — Sometimes. Particularly if there is already a legal proceeding and somebody is seeking access to the AI records. The judge could examine the material in camera and decide what, if anything, loses its privilege.
Your bench-trial objection is excellent. If that same judge eventually determines guilt, we’ve created a problem: you can’t unring a bell. The judge might read extraordinarily prejudicial material, rule it inadmissible, and then supposedly erase it mentally while deciding guilt.
So in a bench trial, I’d prefer a different judge or independent special master to make the privilege determination.
4. Reveal to the prosecution — Only after the independent process determines that a narrowly defined exception applies. And even then, the prosecution shouldn’t necessarily receive the user’s entire AI history. If three paragraphs establish an imminent threat, disclose those three paragraphs—not six years of conversations about his marriage, finances, fantasies and hemorrhoids.
That’s an important principle:
An exception to privilege should pierce only the necessary portion of the privilege, not destroy the privilege altogether.
5. Reveal to the public — Almost never.
This one troubles me most. Even information legitimately disclosed to police or prosecutors needn’t automatically become public property. Suppose somebody tells an AI something intensely embarrassing that becomes tangentially relevant to a prosecution. There is no obvious social benefit in giving newspapers, neighbors, employers and Google access to it forever.
I’d favor sealing AI transcripts by default, with a very high threshold for public release.
And this gets us back to your observation about the difference between AI and your friend. Your friend can testify: “Rodger told me he hated Harry.” You respond: “I never said that.”
A jury evaluates both people. But an authenticated AI transcript can say: August 24, 2026, 3:17:42 PM, USER: “I hate Harry…” And perhaps supply the preceding 47 conversations too.
That isn’t merely another witness. It’s something close to an involuntary diary with perfect recall. Which suggests a useful legal analogy. We shouldn’t think primarily of AI as a person you’ve been talking to.
For privilege purposes, perhaps we should think of the AI conversation as your private papers—an externalized record of your thinking.
That changes the intuition dramatically. The government saying “We want to question your friend about what you told him” feels fairly ordinary. The government saying: “We want every private thought you chose to put into your lifelong electronic diary” feels very different.
And your hierarchy gives us a sensible principle: User → user’s lawyer → independent confidential reviewer → judge → prosecution → public. At every step to the right, the burden for disclosure should become substantially higher.
One correction to our earlier discussion follows from this too. Even when there is a credible imminent threat, the objective isn’t necessarily to give the prosecution evidence. It’s to prevent the threatened harm. Those are different governmental interests.
Police might need “Harry is in immediate danger at this address tonight.” They don’t necessarily need the user’s entire transcript for subsequent prosecution. That’s a distinction I’d build into the law from the beginning: Disclosure necessary to save someone should not automatically become disclosure available to convict someone.
That would make the privilege considerably more meaningful.
RMM: Basically, we should lean toward “innocent until absolutely proven guilty.” In my mind (and in the law), convicting one innocent person is many orders of magnitude worse than freeing a guilty person. That is why we usually require unanimous juries.
Unfortunately, though juries are given instructions, individual jurors tend to think in “preponderance of the evidence,” terms, and if they don’t like the guy, they will vote to convict — thus the unanimity requirement — which is being eaten away in some cases.
Chat GPT: I agree with the direction, but I’d separate two ideas that are getting mixed together. The criminal-law principle isn’t quite “innocent until absolutely proven guilty.” It’s presumed innocent until proven guilty beyond a reasonable doubt. “Absolutely proven” would make conviction nearly impossible; virtually every factual conclusion retains some conceivable possibility of error.
But your underlying value judgment is exactly why criminal law deliberately tilts the scales. Blackstone’s famous formulation was: “It is better that ten guilty persons escape than that one innocent suffer.” And Benjamin Franklin went much further: better that 100 guilty persons escape than one innocent person suffer. The numbers aren’t mathematics; they express the asymmetry you’re talking about.
There is one factual point I’d correct, though. For serious criminal convictions in U.S. state and federal courts, jury unanimity isn’t being eaten away—it was strengthened constitutionally. In Ramos v. Louisiana (2020), the Supreme Court held that the Sixth Amendment requires a unanimous jury verdict to convict a defendant of a serious offense in state court as well as federal court.
Where I strongly agree with your concern is the psychology of juries. The instruction may say beyond a reasonable doubt, but there’s no meter inside twelve people’s heads showing whether somebody has reached 51%, 75%, 90%, or 99%. And things that legally shouldn’t establish guilt—I don’t like him; he looks shifty; why didn’t he testify?; he’s obviously some kind of bastard—can affect human judgment.
That’s one reason the rules surrounding evidence matter so much. We don’t merely tell jurors to ignore unfairly prejudicial information; often we prevent them from hearing it in the first place. Once you’ve heard something, “disregard that statement” asks the brain to perform a trick it isn’t particularly good at.
And that circles beautifully back to your AI-privilege problem. Suppose an AI transcript contains: “I’ve always hated Harry. I’ve sometimes wished the son of a bitch were dead.” Then Harry gets murdered.
That statement could be devastating before a jury—even though millions of perfectly innocent people have wished unpleasant things about somebody who subsequently didn’t happen to be murdered. So the presumption you’re proposing for AI records shouldn’t really be: “Keep it secret because the defendant probably is innocent.”
It should be stronger: “Keep it secret unless the government establishes a compelling legal reason why this particular private communication should lose its protection.”
The burden belongs on the party seeking disclosure. And your independent-arbitrator idea becomes more attractive because of exactly the psychological problem you’ve identified. Don’t ask the jury to unsee something. Don’t let them see it unless they’ve earned the right to see it under the rules of evidence.
I’d extend that even to prosecutors during the initial privilege determination when feasible. A prosecutor who has read 15 years of someone’s innermost AI conversations cannot unread them either. So perhaps the architecture is:
AI conversation starts privileged → independent judicial reviewer examines disputed material privately → only material meeting a narrowly defined exception crosses the wall → admissibility is considered separately → jury sees it only if it passes that second test.
That creates deliberate friction. And in this area, friction is a feature, not a bug.
Because the technological change you’ve identified is extraordinary: for most of human history, our ugliest passing thoughts conveniently disappeared. Now we can accidentally manufacture a searchable, permanent, verbatim archive of them.
The evidentiary system shouldn’t casually convert perfect digital memory into perfect government surveillance.
RMM: Finally, we get to the problem of fakes. Can an AI conversation, whether audio or video. be faked? If so, there should be electronic guardrails against that.
Absolutely. An AI conversation can be faked. And this creates a second problem almost opposite to the one we were just discussing.
A screenshot of an AI conversation is trivially fakeable. An exported-looking transcript can be fabricated. Audio can be synthesized. Video can be deepfaked. As these technologies improve, “Look, here is the recording” increasingly will cease to mean “therefore it happened.”
But your second thought is important: the same technology that creates this problem can provide much stronger authentication than we have had historically. The key isn’t making fakes impossible. That’s probably hopeless. The key is making authentic conversations independently verifiable.
Imagine that every genuine AI conversation quietly received something analogous to a digital fingerprint—a cryptographic signature. It could establish things such as:
This conversation really occurred on this AI service.
These words actually were entered by the user.
These words actually were generated by the AI.
They occurred in this sequence.
The record hasn’t been altered afterward.
Change one character and the authentication fails. That technology essentially exists already in the form of cryptographic hashes and digital signatures. The hard part is designing the legal and privacy infrastructure around it.
And there’s an extremely important wrinkle given everything we’ve just discussed: Authentication must not become surveillance.
We don’t want the solution to fake AI conversations to be a gigantic central database saying: “Here is every conversation Rodger has ever had, permanently stored and available for verification.” That would “solve” forgery by creating the 1984 system we’re trying to avoid.
Instead, I’d want something closer to a user-controlled authenticity certificate. The service cryptographically signs the conversation when it occurs. Later, if you choose to produce that conversation, an independent system can answer:
AUTHENTIC — this matches the originally signed record. or NOT AUTHENTICATED — we cannot establish that this is an original unaltered record. Importantly, the verifier shouldn’t have to receive all your other conversations.
There’s an even nastier problem: selective quotation: Suppose the authentic conversation is: “Sometimes I get so angry at Smith that I think I could kill him. Of course I never would. He’s an irritating bastard, but I wouldn’t hurt him.” Someone produces only: “I think I could kill him.” Every word is genuine. The evidence is nevertheless profoundly misleading.
So your electronic guardrails need to authenticate not only content, but context. A system should be capable of establishing: “Yes, those words are genuine, but they have been extracted from a larger authenticated conversation.”
That doesn’t mean the whole private conversation necessarily gets disclosed. Our independent reviewer could inspect enough surrounding material to determine whether the excerpt fairly represents what was said. And now we’ve developed two complementary legal principles:
Privacy: Nobody gets my AI conversations merely because they might contain something useful.
Authenticity: If an AI conversation is introduced as evidence, nobody gets to manufacture, alter, or misleadingly excerpt it and call it mine.
There’s a third problem I’d add: identity. Even if we prove that the words genuinely were entered into your ChatGPT account, have we proved you typed them? Maybe your grandson borrowed your laptop. Someone stole your password. You left the computer open. Someone deliberately typed something incriminating into your account. So authentication has at least three separate questions:
1. Is this an authentic record from the AI system?
2. Has the record been altered or selectively presented?
3. Did the accused person actually create the user-side statements?
Number 3 is much harder.
And that’s another reason I wouldn’t allow “authenticated AI transcript = proven confession.” It should be evidence whose provenance can be established, not magical evidence that automatically proves authorship or guilt. Your original observation about he said/she said therefore has acquired an interesting qualification.
AI potentially eliminates: “Did these words ever appear in the conversation?” It does not necessarily eliminate: “Who entered them, what did they mean, and what was their context?”
Those distinctions should be built into any eventual AI-evidence law. The guiding principle I’d use is wonderfully old-fashioned: Technology should make evidence easier to authenticate—not easier to believe. A jury still has to decide what authenticated evidence means.
Ha! Yes. O.J. could derail our entire theory of evidence before dinner. And actually, his case illustrates your point rather nicely: even extremely powerful evidence doesn’t determine what a jury will do with it. Authentication, admissibility, reasonable doubt, juror psychology, credibility of police and prosecutors, and simple human emotion all intervene between evidence and verdict.
So we’ll leave O.J. peacefully parked beside the bloody glove and continue designing civilization.
Rodger Malcolm Mitchell
Source: https://mythfighter.com/2026/08/24/should-ai-convict-or-free-the-accused/
Anyone can join.
Anyone can contribute.
Anyone can become informed about their world.
"United We Stand" Click Here To Create Your Personal Citizen Journalist Account Today, Be Sure To Invite Your Friends.
Before It’s News® is a community of individuals who report on what’s going on around them, from all around the world. Anyone can join. Anyone can contribute. Anyone can become informed about their world. "United We Stand" Click Here To Create Your Personal Citizen Journalist Account Today, Be Sure To Invite Your Friends.
LION'S MANE PRODUCT
Try Our Lion’s Mane WHOLE MIND Nootropic Blend 60 Capsules
Mushrooms are having a moment. One fabulous fungus in particular, lion’s mane, may help improve memory, depression and anxiety symptoms. They are also an excellent source of nutrients that show promise as a therapy for dementia, and other neurodegenerative diseases. If you’re living with anxiety or depression, you may be curious about all the therapy options out there — including the natural ones.Our Lion’s Mane WHOLE MIND Nootropic Blend has been formulated to utilize the potency of Lion’s mane but also include the benefits of four other Highly Beneficial Mushrooms. Synergistically, they work together to Build your health through improving cognitive function and immunity regardless of your age. Our Nootropic not only improves your Cognitive Function and Activates your Immune System, but it benefits growth of Essential Gut Flora, further enhancing your Vitality.
Our Formula includes: Lion’s Mane Mushrooms which Increase Brain Power through nerve growth, lessen anxiety, reduce depression, and improve concentration. Its an excellent adaptogen, promotes sleep and improves immunity. Shiitake Mushrooms which Fight cancer cells and infectious disease, boost the immune system, promotes brain function, and serves as a source of B vitamins. Maitake Mushrooms which regulate blood sugar levels of diabetics, reduce hypertension and boosts the immune system. Reishi Mushrooms which Fight inflammation, liver disease, fatigue, tumor growth and cancer. They Improve skin disorders and soothes digestive problems, stomach ulcers and leaky gut syndrome. Chaga Mushrooms which have anti-aging effects, boost immune function, improve stamina and athletic performance, even act as a natural aphrodisiac, fighting diabetes and improving liver function. Try Our Lion’s Mane WHOLE MIND Nootropic Blend 60 Capsules Today. Be 100% Satisfied or Receive a Full Money Back Guarantee. Order Yours Today by Following This Link.
| Visits: | 1,830,006,069 |
| Stories: | 8,692,764 |
Whistler Blowers, Insiders
