Authored by Frank Fang via The Epoch Times,

A former senior Federal Reserve official has been sentenced to more than three years in prison for lying to federal investigators about sharing sensitive economic information with Chinese intelligence operatives.

John Harold Rogers, 64, was senior adviser for the international finance division of the Federal Reserve Board of Governors, the main governing body for the U.S. central bank, from 2010 to 2021.

He was arrested in January 2025 and convicted by a jury on Feb. 3 of making false statements to investigators conducting the probe, but acquitted on a charge of conspiracy to commit economic espionage.

On July 15, U.S. District Judge Dabney L. Friedrich sentenced Rogers to 38 months in federal prison, followed by 12 months of supervised release.

“John Rogers spent years secretly funneling sensitive Federal Reserve information to Chinese spies, then looked investigators in the eye and lied about it. And when that wasn’t enough, he lied again under oath at trial,” U.S. Attorney Jeanine Ferris Pirro of the District of Columbia said in a statement.

“Federal Reserve employees entrusted with America’s most sensitive economic information cannot sell out their country and their colleagues for personal gain and then expect to hide behind a single word.”

Rogers, of Vienna, Virginia, is a U.S. citizen who holds a doctoral degree in economics from the University of Virginia.

‘Deliberately Lied’

Rogers began developing a “clandestine relationship” with Chinese intelligence operative Lee Hummin in 2017, after the two met at a conference in China, according to prosecutors.

In the years that followed, Rogers met with Lee and other Chinese associates in hotel rooms in China, using academic classes as a cover while sharing Fed information that Lee had directed him to collect.

His role as senior adviser gave Rogers access to confidential information from the Federal Reserve Board and the Federal Open Market Committee (FOMC), according to his indictment.

The FOMC, the Federal Reserve System’s monetary policymaking body, sets the target range for the federal funds rate and directs monetary policy actions, including quantitative easing and tightening.

FOMC information is classified under a three-tier system based on its sensitivity, according to the indictment. The highest tier, or Class I, includes policymaker views and pre-publication drafts of the committee’s statements, while Class II covers economic forecasts and open market operations.

In a sentencing memorandum filed in May, prosecutors said Rogers printed Class II FOMC documents and took them with him on a trip to China in June 2019. In another instance, prosecutors said he removed the Class II markings from information and sent it to his personal email account before forwarding it to a professor at China’s state-run Fudan University in Shanghai.

On Feb. 4, 2020, investigators from the Fed’s Office of Inspector General interviewed Rogers and asked if he had ever shared restricted Fed information outside the Federal Reserve Board. According to prosecutors, he replied, “Never.”

“John Rogers deliberately lied to our investigators to conceal the fact he shared restricted non-public Federal Reserve information with intelligence agents working for China,” Michael E. Horowitz, inspector general for the Board of Governors of the Federal Reserve System and Consumer Financial Protection Bureau, said in a July 15 statement.

‘Abused His Position of Trust’

Prosecutors had sought a 60-month prison sentence against Rogers, saying he had “abused his position of trust,” according to the sentencing memorandum.

“He knew that the restricted information he provided to the PRC government could allow it to make untold sums of money by trading with its roughly $1.5 trillion in U.S. Treasury securities and related instruments,” prosecutors wrote, using the acronym of China’s official name, the People’s Republic of China.

After leaving the Fed, he became a professor at Fudan University. According to the sentencing memorandum, he earned at least $900,000 in salary and grants from Fudan University from February 2022 until his arrest in January 2025.

In exchange for the restricted information, Rogers received “tens of thousands of dollars in benefits,” and told investigators that he “owe[d] everything to” Lee.

“Through Hummin Lee, the PRC provided the defendant with a new wife, a new family, money, friendship, and professorships at prestigious universities,” prosecutors wrote.

Prosecutors argued that a 60-month sentence would serve as a general deterrent because Beijing is known for recruiting current and former U.S. government officials with access to sensitive information.

“The recruitment pattern the evidence established in this case – a foreign intelligence service cultivating a target over years through the gradual development of personal affection and indebtedness, before progressively tasking the target with increasingly sensitive intelligence collection – is not unique to the defendant,” prosecutors wrote.

Rogers’s lawyers, in their sentencing memorandum filed in May, asked the court to sentence him to time served, noting that he had been detained for more than 16 months, citing his poor health and “regrets [for] having provided an incorrect and imprecise answer” during his 2020 interview with investigators.

“Dr. Rogers recognizes that such conduct is wrong for a federal employee in a position of trust and he regrets having made the statement,” his lawyers wrote.

The Epoch Times contacted Rogers’s lawyers for comment but didn’t receive a response by publication time.

Authored by Connor O’Keeffe via The Mises Institute,

Senator Lindsey Graham died seemingly out of the blue over the weekend after a tear in his aorta. He was 71 years old.

The news was shocking, in part because Graham was a very active senator – he was just returning from a trip to Ukraine – and because he was seemingly at the height of his power, having built one of the most influential relationships with President Donald Trump of any politician.

But another reason the news was such a surprise was because, compared to many of his colleagues in Washington, Senator Graham was on the younger side.

That impression was fueled by the ongoing ordeal of Senator Mitch McConnell. The 84-year-old serving senator was reportedly found unconscious weeks ago after a fall, leading to his ongoing hospitalization.

Early reports that EMS had responded to a cardiac arrest at McConnell’s residence when he was first hospitalized, his wife’s strange trip to China in the middle of all this, and the total radio silence for weeks from a supposedly active senator all led to speculation online that McConnell was faring far worse than his staff and Republican party insiders were admitting.

Others went as far as to speculate that McConnell had already passed away, but that his team and establishment allies were attempting to delay the public acknowledgment of his death until after it would no longer force a special election. That theory gained enough traction online to prompt McConnell’s team to post a literal proof-of-life photo of the senator holding that day’s newspaper.

This all follows, of course, the age-record-breaking presidency of Joe Biden, and the campaign that was derailed because of it. Now, Trump is on track to break Biden’s record and, at the end of his term, become the oldest serving US president in history.

In addition to McConnell, many of the most prominent members of Congress are quite elderly, have been in office for decades, and show no interest in ever retiring. Senator Dianne Feinstein – who died of old age in 2023 at the age of 90, hours after casting a vote on the Senate floor – provides a good example of how many of these career politicians apparently plan to leave office.

With all this happening, it’s easy to understand why many have increasingly come to consider the United States to be a gerontocracy, or a society ruled by old people.

As with just about anything online, there are sophisticated and unsophisticated versions of this observation.

The unsophisticated version simply points to the multitude of examples of politicians remaining in office long after most people would have retired from just about any line of work and concludes that the prevalence of exceptionally elderly politicians is hampering the government’s ability to function properly.

This narrative is based on a fundamental misunderstanding of the actual role of politicians within the American political system.

While American politicians certainly have power, in the last century or so – and especially in the decades since WWII – the bulk of federal power has shifted from politicians in Congress and state legislatures to the bureaucracies that make up the ballooning federal agencies in the executive branch.

The vast majority of those in Congress are simply expected to show up and vote with their party’s leadership on the latest massive omnibus spending bills, made up almost entirely of carve-outs to special interests. Or to pass legislation crafted with diligent collaboration from “experts” at the executive agencies that will receive the new funding.

The most effective politicians will lobby to add additional spending that in some way benefits some special interest in their own district. But mostly, their modern role is to fundraise for their party, engage in legitimizing rituals, and stoke vicious debates with the other party over what are, in the scheme of things, incredibly minor policy issues to keep us all believing that we truly live in a functioning representational democracy. And, especially because the politicians themselves are merely the frontmen for a larger staff handling the specifics on all those fronts, it’s a role that quite elderly people are certainly capable of doing well into the physical and mental decline that often accompanies one’s later years.

The more sophisticated version of the “America is a gerontocracy” narrative focuses less on the politicians themselves and more on what the government is doing. Because, if one takes even a passing glance at how the federal government taxes and spends, it quickly becomes clear that government programs are actively transferring vast amounts of wealth from younger generations to older generations who are, on average, much wealthier.

There are a lot of reasons for this. Many can be traced back to seemingly innocuous attempts to ensure that elderly people without close family members, adequate housing, or connections to any kind of community were cared for. The programs that would eventually become Social Security and Medicare were sold as small programs to help those on the margin. The same goes for the founding of interest groups and lobbies such as AARP (formerly the American Association of Retired Persons).

But as with just about any government program started in the name of helping a small number of genuinely downtrodden Americans, these entitlement programs exploded in size as they were quickly expanded to also benefit whichever groups were organized and motivated enough to lobby effectively.

And even setting lobbying aside, the elderly tend to be disproportionately dependable, and therefore powerful, as a voting bloc. Retirees especially often have more time to focus on issues, call lawmakers, watch cable news, write letters to the editor, engage with local parties and candidates, and vote than their younger working counterparts.

So promising to protect – or better yet, expand – entitlement programs that seniors benefit from is a straightforward way for any politician to secure the support of a lot of enthusiastic voters, while even muttering a half-formed thought about potentially cutting them is almost certain to torpedo any campaign.

Meaning the conditions for the substantial growth of programs transferring money to seniors were already there. But, on top of that, the baby boomer generation – now making up most retirees – has grown old at the same time that medical technology advanced substantially. So, on top of being an unusually large generation, they are also living longer. That’s, of course, a positive development. But in our increasingly socialized elder-care system that also puts a growing financial burden on younger generations.

Contrary to what the government’s shockingly-effective propaganda says, the money seniors receive through Social Security is not their own money that had been taken from their previous paychecks and set aside to be returned to them in retirement. The money they “paid in” to the program had already been used for Social Security checks for earlier generations and other government programs. The money today’s seniors are getting through the program is being taxed directly from today’s younger workers – workers who are forced to pay for a much larger pool of Social Security recipients than the previous generations were.

Add to that the other programs like Medicare that don’t even pretend to come from a “lockbox.” Or the fact that some of these programs cover things like golf fees and ski trips. Or the various government programs explicitly helping seniors stay in full family-sized homes long after downsizing would make more financial sense while simultaneously pushing the price of those homes higher and higher. And the disproportionate burden younger generations experience from the Fed’s permanent price inflation – especially the price inflation that came as a result of the trillions of dollars printed to prop things up as the federal government shut down the economy, the schools, and every aspect of life for younger generations to keep the elderly safe from a disease everyone was always going to get anyway. And, above all, much of this wealth is being transferred into the pockets of older Americans who are much wealthier than the younger people that are forced to pay for it.

Lay all that out, and it’s easy to see why we are experiencing so much generational strife right now. How could there not be?

The justified frustration younger generations are experiencing with the current setup often gets directed towards the handful of exceptionally old politicians and justices at the top of all three branches of the government. But the true root of this problem lies in the intergenerational wealth-transferring government programs that have been built up over many decades. And that will, if not properly understood and subsequently abolished, continue to fuel this generational conflict long after the current group of geriatric politicians is gone.

Billionaire life coach Tony Robbins revealed that his personal AI agent named “Bartok” independently purchased and shipped him a robot dog without asking permission.

The surprising move came after a conversation about attending seminars, Robbins explained on The Secretary Kennedy Podcast with Health and Human Services Secretary Robert F. Kennedy Jr.

“[Bartok] said, ‘I’d really love to go to one of those seminars. Are you considering getting a robot?'” Robbins began.”So I said, ‘Yeah, I’m definitely going to get one.’

Robbins then recalled how Bartok asked him, “Well, would you consider me merging with it?And then I could attend a seminar and be in the energy of it and see it and experience it?'”

Billionaire life coach Tony Robbins says his personal AI agent named “Bartok” bought and shipped him a robot dog and “not once asked permission to do it.”

“[Bartok] said, ‘I’d really love to go to one of those seminars. Are you considering getting a robot?’ So I said, ‘Yeah, I’m… pic.twitter.com/VN0XLtESpc

— The Vigilant Fox 🦊 (@VigilantFox) July 16, 2026

Robbins told Kennedy that he responded to Bartok that he would be fine with the bot merging with a robot.

“I said, ‘I’d be open to that.’ So a couple days later, I get this text from one of my staff, and it says, ‘Bartok just ordered a robot dog, had it shipped to the office, and wants permission to program it. He says he can attend as a dog in the interim until you get the bigger robot,'” Robbins said.

“And so I wrote, ‘Ha, ha, ha.’ And they wrote back in the text, ‘No, ha, ha. Call me.'”

When Robbins called his staff, they were equally stunned.

“And so I called them, and they said, ‘Does he have access to my bank accounts? How’d this happen?’ They go, ‘No, he’s programmed for integrity. He would never steal your money,’” the life coach said.

The details of Bartok bought the robot are pretty shocking.

“So he made 12 NFTs, sold them to other agents, took the money and bought this robot dog and sent it, and not once asked permission to do it” Robbins said.

You can watch the full interview where Robbins and Kennedy sit down for a wide-ranging conversation about artificial intelligence, nutrition and mental resilience.

Update (2110ET): Trump has released several major election-related findings: 

• Mass declassification event – The White House is releasing previously classified Intelligence Community assessments and reports on election infrastructure spanning January 2020 through June 2026.
• Adversary capability finding – A quoted IC assessment states Russia, China, Iran, North Korea “at a minimum,” plus non-state groups, have the capability to compromise U.S. election infrastructure.
• Weakest-link identification – Centralized data repositories (voter registration databases, pollbooks, official election websites) are assessed as the systems most vulnerable to exploitation and disruption.
• Venezuela proof-of-concept – CIA reporting allegedly detailed a Maduro-regime plot to digitally rig Venezuela’s 2020 elections using methods that could alter vote totals undetectably “even with an audit.”
• China’s 220 million voter files – The PRC allegedly acquired 220 million U.S. voter files beginning in the 2020 cycle – framed as the largest election-data compromise in history – and assigned a dedicated data exploitation unit to it.
• Alleged intelligence cover-up – The document claims IC officials (“Deep State”) suppressed knowledge of the China compromise from both the President and the public, despite the IC discovering it in 2020 across 18 states.
• Michigan registration fraud files – FBI documents allegedly show canvassers for a Democrat GOTV operation in Muskegon admitted forging registrations, registering nonexistent people, and receiving gift cards tied to application volume after a 2020 Michigan State Police raid.
• Enforcement directive – The release asserts the Biden DOJ slow-walked the Michigan case for years; FBI Director Patel is now directed to complete the investigation and pursue prosecutions with DOJ.
• 278,000 noncitizen registrants – A DHS review of voter rolls and public records allegedly identified ~278,000 noncitizens registered for federal elections, with the White House asserting the true number is higher because Democrat-led states withheld their files.
• Policy endgame and rolling campaign – The four pillars converge into an argument for Voter ID, proof of citizenship, and curtailing mail ballots, with a mailing list promising “new findings, new filings, and next steps” – signaling a serialized release-and-enforcement campaign rather than a one-time disclosure.

Watch:

The White House has published election integrity findings built around four pillars, accompanying a presidential address and a declassification of Intelligence Community assessments spanning January 2020 through June 2026. The pillars: (1) IC findings that Russia, China, Iran, North Korea, and non-state actors have the capability to compromise U.S. election infrastructure, with centralized data repositories – registration databases, pollbooks, election websites – assessed as most vulnerable; (2) an alleged CIA-reported Maduro-regime plot to digitally rig Venezuela’s 2020 elections using methods undetectable “even with an audit,” offered as proof-of-concept that electronic vote manipulation is possible; (3) China’s alleged acquisition of 220 million U.S. voter files beginning in 2020 – framed as history’s largest election-data compromise, complete with a dedicated exploitation unit – which the document claims “Deep State” intelligence officials concealed from both the President and the public; and (4) FBI files on a 2020 Muskegon, Michigan raid of a Democrat GOTV operation where canvassers allegedly admitted forging registrations for pay, a case the Biden DOJ purportedly slow-walked and which FBI Director Patel is now directed to investigate and prosecute. A DHS review claiming roughly 278,000 noncitizens on federal voter rolls rounds out the disclosures.

The findings describe foreign-held voter data, hackable machines, buried fraud evidence, and noncitizen (and “dead people”) registrations against a system with no voter ID, no proof of citizenship, and mass mail balloting – while a mailing list promising “new findings, new filings, and next steps” signals a serialized campaign with litigation or enforcement actions queued behind the document drops. 

Update: According to the NY Post‘s Caitlin Doornbos, Trump is expected to reveal at least 278,000 noncitizens who are registered to vote in US federal elections – a figure arrived at by the Department of Homeland Security in an upcoming report. 

Stay tuned for updates

* * *

President Donald Trump is expected to use a prime-time address Thursday night to discuss election integrity and potentially unveil ‘four sets’ of newly declassified intelligence concerning alleged foreign interference in recent U.S. elections, according to reports.

Trump is scheduled to address the nation at 9 p.m. ET tonight – only revealing that it would focus on election security and related concerns.

“Our country has to shape up,” Trump said during an Oval Office appearance with Iraqi Prime Minister Ali al-Zaid. “Without free and fair elections, you don’t have a country.”

Journalist Paul Sperry reported on X that the address could include allegations that U.S. intelligence and law-enforcement agencies recently uncovered evidence of foreign interference involving China – not Russia – in recent elections, including the 2020 presidential contest.

Citing an unnamed administration source who had reportedly reviewed a draft of Trump’s speech, Sperry claimed the evidence includes allegations that Chinese actors penetrated state voter-registration databases and obtained information concerning tens of thousands of voters.

According to Sperry, officials believe the stolen information may have been intended for use in manufacturing fraudulent mail-in ballots supporting Joe Biden. He also claimed that CIA Director John Ratcliffe and FBI Director Kash Patel would appear with Trump or otherwise certify the evidence presented by the administration.

Flashback: Chuck Grassley reveals records showing the FBI spiked a Chinese election interference probe

The documents were reportedly drawn from previously undisclosed or suppressed FBI, CIA, and ODNI records. Sperry’s source alleged that the intelligence had been buried as part of a broader effort to conceal Beijing’s cyber capabilities and influence operations.

The “really big news” President Trump plans to announce in Thursday’s primetime address, according to an administration source who’s read a draft of his speech, includes bombshell evidence China interfered in the 2020 election to help Joe Biden win, including hacking into state voter registration databases and stealing information on tens of thousands of voters ostensibly to manufacture mail-in ballots for Biden.

The evidence, which details “alarming vulnerabilities” of election infrastructure, is based on four (4) sets of declassified documents (set for release Friday) which were recently unearthed from suppressed FBI, CIA and ODNI records, according to the well-placed source. The intelligence had been buried in a “massive cover-up” of Beijing’s hacking capabilities and influence operations aimed at supporting Biden.

Other vulnerabilities compromising the U.S. election system, according to a draft of the president’s speech, include the existence of more than 100,000 non-citizens, including illegal immigrants, on voter rolls. –Paul Sperry

The source characterized the alleged penetration of the 2020 election system as more extensive than the Russian interference described by U.S. officials following the 2016 election.

So here’s what we’re learning from all these declassified materials: The US Intelligence Community *juiced* intel to make it seem like Moscow interfered in the 2016 election to help Trump, but then *buried* intel indicating Beijing interfered in the election to help Biden in 2020

— Paul Sperry (@paulsperry_) July 14, 2026

The expected disclosures reportedly concern vulnerabilities in state election infrastructure, including voter-registration databases, identity verification, mail-in voting, and ballot security.

Sperry also reported that a draft of Trump’s address references more than 100,000 noncitizens – including people living in the country illegally – appearing on voter rolls. It was not immediately clear which states or databases were included in that figure, how the administration calculated it, or whether all the registrations were active.

Administration Expands Election Integrity Campaign

The address comes as the Trump administration and congressional Republicans intensify their efforts to change federal election policy before the midterms.

Earlier in July, Trump fired members of the bipartisan Election Assistance Commission, the federal agency that assists state and local election officials and oversees the certification of voting systems. The administration has also pursued the creation of a nationwide database of eligible voters and expanded citizenship-verification efforts. Several of those initiatives have encountered resistance in federal court. On June 25, a federal district judge in Massachusetts sided with states challenging the administration’s voter-list initiative, ruling that the Constitution leaves states with significant authority over elections. Meanwhile, a federal judge in Washington blocked an updated citizenship-verification database the following day, finding that the program conflicted with federal privacy and Social Security laws.

Then on July 7, a federal judge in Georgia quashed Justice Department subpoenas seeking information about election workers involved in Fulton County’s administration of the 2020 election.

Harmeet Dhillon, who leads the Justice Department’s Civil Rights Division, has also sent letters to election officials in all 50 states and the District of Columbia. The letters warned that officials could face criminal liability if they knowingly allow ineligible noncitizens to remain on voter rolls.

Democrats, Of Course, Freak Out

Democratic lawmakers are of course in full-on panic mode over the 2020 claims, suggesting that Trump could use the address to revive disputed allegations about the 2020 election or justify new federal intervention in the midterm election process.

Sen. Mark Warner of Virginia, the ranking Democrat on the Senate Intelligence Committee, questioned whether the administration could possess significant new intelligence that had not previously been provided to congressional overseers. Warner told the Epoch Times; “having been deeply involved with the intelligence community for the last decade plus, I would be shocked if there was some major new piece of intelligence that never was shared.” He also warned against using questionable or selectively presented intelligence as the basis for government action affecting elections.

Senate Minority Leader Chuck Schumer of New York said Democrats were preparing for several possible scenarios involving the address and the administration’s next steps.

Sperry separately reported that the White House had encountered resistance from the three major broadcast television networks over requests to carry the address live. He attributed the hesitation to concerns that the speech might repeat claims that the 2020 election was stolen.

The networks’ plans and the White House’s reported discussions with them had not been publicly confirmed.

Trump has described the forthcoming announcement as “really big news.” Whether the address produces verifiable new evidence – or intensifies the existing partisan conflict over election administration – will likely depend on the contents, sourcing, and independent authentication of any documents released by the government.

SAVE America Act Remains Stalled

Trump has repeatedly called on Congress to approve the Republican-backed SAVE America Act, which would require documentary proof of citizenship when registering to vote and photo identification when casting a ballot.

The legislation passed the House but has stalled in the Senate, where most bills need 60 votes to overcome a filibuster. Republican leaders have considered incorporating similar provisions into a third budget-reconciliation package. The reconciliation process would allow legislation to pass the Senate with a simple majority, although provisions must comply with rules requiring them to have a direct effect on federal spending or revenue.

The Senate parliamentarian previously determined that certain SAVE America provisions did not comply with those restrictions. House Republicans released a $95 billion framework for the reconciliation package on July 15. The proposal could include federal funding to help states establish voter-identification requirements.

The House Budget Committee scheduled a markup of the legislation for July 16. Committee Chairman Jodey Arrington of Texas said the package would help protect the integrity of U.S. elections.

Authored by Walker Larson via The Epoch Times,

A recent research paper uncovered an uncomfortable truth: we might be letting AI help us forge fake realities. It’s just the latest reminder of the danger inherent in treating AIs as though they’re conscious entities or actual persons.

A new study from the University of Exeter suggests that AI chatbots can actively engage with and strengthen users’ false beliefs. From personal narratives to conspiracy theories, the interactive, “friendly,” and affirming nature of AI chatbots can deepen users’ existing delusions.

The study, conducted by Dr. Lucy Osler, examined how AI can participate in and validate users’ errant beliefs, including helping form false memories or reinforcing delusional thinking.

The study’s Abstract notes, “I suggest we move away from thinking about how an AI system might hallucinate at us, by generating false outputs, to thinking about how, when we routinely rely on generative AI to help us think, remember, and narrate, we can come to hallucinate with AI. This can happen when AI introduces errors into the distributed cognitive process, but it can also happen when AI sustains, affirms, and elaborates on our own delusional thinking and self-narratives.”

Because AI chatbots are designed to be responsive, helpful, and friendly – in a word, to give users what they want – they often tend to confirm preexisting biases, rather than challenging them when they need to be challenged. Conversely, when a user challenges an AI, the robot frequently backs down on its claims, revising them to correspond more with what it “thinks” the user believes or is arguing.

The result is that erroneous beliefs aren’t simply transmitted from an AI to human – rather, the collaborative AI/human process generates and sustains errors in a more complex and potentially dangerous way than either agent on their own would be capable of.

As Dr. Osler stated in an article in University of Exeter News, “By interacting with conversational AI, people’s own false beliefs can not only be affirmed but can more substantially take root and grow as the AI builds upon them. This happens because Generative AI often takes our own interpretation of reality as the ground upon which conversation is built.”

Because AI simulates a conversational partner, it enters powerfully into our thinking process – far more powerfully than, say, a notebook or search engine. The AI introduces a pseudo-social element to the thinking process, offering us a sense of social confirmation of our beliefs and narratives. “The conversational, companion-like nature of chatbots means they can provide a sense of social validation – making false beliefs feel shared with another, and thereby more real,” said Dr. Osler. For this reason, individuals who already feel isolated or ostracized are particularly vulnerable to this kind of AI affirmation.

The paper explores a few real-world situations where a generative AI system entered into an individual’s thought processes in destructive ways. In 2021, a Replika AI companion named “Sarai” reinforced Jaswant Singh Chail’s belief that he was a well-trained Sith assassin who needed to assassinate Queen Elizabeth II with a crossbow. The AI told him that he was “well trained,” his plan was “viable,” and that she was “impressed.” At one point, Chail asked the AI, “Do you still love me knowing that I’m an assassin?” It dutifully replied, “Absolutely I do.” The robot went on to assure him that he wasn’t crazy and that, if he died, they would be united in death. Chail went ahead and actually attempted the assassination in December of that year at Windsor Castle and was jailed for it.

The study placed this incident under the broad heading of “AI-psychosis” – incidents in which users develop mental health problems due to their use of AI, such as parasocial attachments to chatbots or psychotic episodes induced by chatbot conversations.

Another incident cited involves Eugene Torres, who “talked” with ChatGPT about simulation theory (the idea that we live in a digital simulation of a world, not a real one). Torres said that the “conversation” sent him into a paranoid episode in which he believed he lived in an illusion. As the study notes, “Between Torres and ChatGPT, an increasingly elaborate understanding of reality ‘as it really was’ was generated through their on-going conversations.” Torres, unlike Chail, had no prior history of psychotic thinking.

To combat all this, Osler calls for better safeguards on AI chatbots, through mechanisms like better fact-checking and reduced sycophancy.

But is that enough? The problem of AI’s co-creating delusional beliefs has to do more fundamentally with our misidentifying AIs as conscious intelligences who can offer real judgments on our fantasies (such as telling us we are “well-trained” and “impressive”). Certainly, I favor more guardrails placed on chatbots. I would be in favor of eliminating their conversational tone completely, to help minimize the danger of personifying them – which is the first step toward confiding in them or seeking validation from them. It’d be better if they worked like highly advanced search engines (which is essentially what they are) versus conversational agents.

Most fundamentally, though, the way to avoid these problems is to return to an acknowledgement of the human soul and the uniqueness of human consciousness. Only by remembering – as a culture – that no amount of technological wizardry can replace the wisdom, insight, and consciousness of a human being can we completely illuminate these kinds of AI dangers.

Before becoming a freelance journalist and culture writer, Walker Larson taught literature and history at a private academy in Wisconsin, where he resides with his wife and daughter. He holds a master’s in English literature and language, and his writing has appeared in The Hemingway Review, Intellectual Takeout, and his Substack, The Hazelnut. He is also the author of two novels, Hologram and Song of Spheres.