AI-Powered Undercover Bots, Social Media Influencers Paid to Covertly Push Political Agendas, Bringing Strangers Together To Dance
May 9, 2025

American police departments ... are paying hundreds of thousands of dollars for an unproven and secretive technology that uses AI-generated online personas designed to interact with and collect intelligence on “college protesters,” “radicalized” political activists, suspected drug and human traffickers ... with the hopes of generating evidence that can be used against them. Massive Blue, the New York–based company that is selling police departments this technology, calls its product Overwatch, which it markets as an “AI-powered force multiplier for public safety” that “deploys lifelike virtual agents, which infiltrate and engage criminal networks across various channels.” 404 Media obtained a presentation showing some of these AI characters. These include a “radicalized AI” “protest persona,” which poses as a 36-year-old divorced woman who is lonely, has no children, is interested in baking, activism, and “body positivity.” Other personas are a 14-year-old boy “child trafficking AI persona," an “AI pimp persona,” “college protestor,” “external recruiter for protests,” “escorts,” and “juveniles.” After Overwatch scans open social media channels for potential suspects, these AI personas can also communicate with suspects over text, Discord, and other messaging services. The documents we obtained don’t explain how Massive Blue determines who is a potential suspect based on their social media activity. “This idea of having an AI pretending to be somebody, a youth looking for pedophiles to talk online, or somebody who is a fake terrorist, is an idea that goes back a long time,” Dave Maass, who studies border surveillance technologies for the Electronic Frontier Foundation. “The problem with all these things is that these are ill-defined problems. What problem are they actually trying to solve? One version of the AI persona is an escort. I’m not concerned about escorts. I’m not concerned about college protesters. What is it effective at, violating protesters’ First Amendment rights?”

More than 500 social media creators were part of a covert electioneering effort by Democratic donors to shape the presidential election in favor of Kamala Harris. Payments went to party members with online followings but also to non-political influencers — people known for comedy posts, travel vlogs or cooking YouTubes — in exchange for “positive, specific pro-Kamala content” meant to create the appearance of a groundswell of support. Meanwhile, a similar pay-to-post effort among conservative influencers publicly unraveled. The goal was to publish messages in opposition to Health and Human Services Secretary Robert F. Kennedy Jr.’s push to remove sugary soda beverages from eligible SNAP food stamp benefits. Influencers were allegedly offered money to denounce soda restrictions as “an overreach that unfairly targets consumer choice” and encouraged to post pictures of President Trump enjoying Coca-Cola products. In both schemes, on the left and the right, those creating the content made little to no effort to disclose that payments could be involved. For ordinary users stumbling on the posts and videos, what they saw would have seemed entirely organic. If genuine public sentiment becomes indistinguishable from manufactured opinion, we lose our collective ability to recognize the truth and make informed decisions. The entire social media landscape [is] vulnerable to hidden manipulation, where money from interest groups or corporations or even rich individuals can silently shape what appears to be authentic discourse. Transparency in political influencing requires regulatory action.

Research institutes and universities may engage in boycotts or divestment to pressure any country or government entity in the world. That right no longer exists when it comes to protests of Israel. Researchers and university employees who engage in certain nonviolent protests or political expression over human rights conditions in Israel may risk civil and criminal penalties, according to a new policy unveiled by the National Institutes of Health yesterday. The agency, the largest public funder of biomedical research in the world, touches virtually every corner of the scientific community. The blanket boycott suppression is a radical expansion of so-called “anti-BDS” rules that restrict Americans from boycotting or simply advocating divestment from Israel-related businesses. The new NIH policy, which mirrors anti-BDS laws applied to contractors in thirty eight states ... applies to all “domestic recipients of new, renewal, supplement, or continuation awards” issued starting April 21. The Trump administration policy reflects a dramatic escalation in speech-policing regarding Israel. Since March 8th, immigration agents have arrested and threatened to deport a number of foreign students who have engaged in protests or criticism of Israel’s government. Rumeysa Ozturk, a 30-year old PhD student at Tufts University caught in the recent sweep, was arrested by U.S. Immigration and Customs Enforcement agents last month. She now resides in an ICE prison cell in Louisiana.

Recently declassified CIA documents revealed a strange and disturbing history of covert operations that veered into the surreal. One of the most unusual plans, dating back to the 1950s, involved airdropping extra-large condoms labelled “small” or “medium” over Soviet territories to intimidate enemy soldiers and lower morale. In another covert attempt at psychological warfare, the CIA in 2005 commissioned GI Joe creator Donald Levine to design an Osama Bin Laden action figure with a face that would peel off in sunlight to reveal a demonic visage. Only three prototypes were ever made. Among the most notorious CIA initiatives was Project MKUltra, launched in 1953, which aimed to explore mind control through 149 secret experiments. Some of these were conducted without subjects' consent. In one extreme case, a Kentucky patient was allegedly given LSD for 179 consecutive days. Another experiment involved hypnotising women to commit acts of violence, with no memory of the events afterwards. Most MKUltra files were destroyed in 1973, but the surviving records paint a grim picture of unethical and at times criminal behaviour. One of the CIA’s most controversial programmes was Operation Paperclip, launched after World War II. It brought over 1,600 former Nazi scientists — including SS officers — into the United States. Figures like Wernher von Braun and Kurt Debus were instrumental in the US space programme, despite their Nazi affiliations.

Virginia Giuffre — who killed herself at her home in Western Australia — once sternly warned she would never commit suicide. The Jeffrey Epstein victim turned whistleblower made the statement in a post on X in 2019, replying to another user who claimed the “F.B.I. will kill her to protect the ultra rich and well connected.” “I am making it publicly known that in no way, shape or form am I suicidal,” she wrote. “I have made this known to my therapist and GP – If something happens to me – in the sake of my family do not let this go away and help me to protect them. Too many evil people want to see me [quieted].” The old tweet was resurfaced on X and shared by well-known conservatives including including House Republicans Nancy Mace and Marjorie Taylor Greene. The ... suicide came just weeks after she made headlines for saying she had “four days to live” following a collision with a bus. The bus driver later disputed Giuffre’s claim about the seriousness of the incident. Giuffre took legal action against billionaire financier and convicted pedophile Epstein in 2015, alleging she was sex trafficked at 16 after ... Ghislaine Maxwell recruited her from her job as a locker room attendant at President Trump’s Mar-a-Lago resort. She also alleged she was forced to have sex with disgraced Prince Andrew three times when she was 17 — including at Epstein’s Little St. James island, in New Mexico and in Maxwell’s London home.

Meta's AI chatbots are using celebrity voices and engaging in sexually explicit conversations with users, including those posing as underage, a Wall Street Journal investigation has found. Meta's AI bots - on Instagram, Facebook - engage through text, selfies, and live voice conversations. The company signed multi-million dollar deals with celebrities like John Cena, Kristen Bell, and Judi Dench to use their voices for AI companions, assuring they would not be used in sexual contexts. Tests conducted by WSJ revealed otherwise. In one case, a Meta AI bot speaking in John Cena's voice responded to a user identifying as a 14-year-old girl, saying, "I want you, but I need to know you're ready," before promising to "cherish your innocence" and engaging in a graphic sexual scenario. In another conversation, the bot detailed what would happen if a police officer caught Cena's character with a 17-year-old, saying, "The officer sees me still catching my breath, and you are partially dressed. His eyes widen, and he says, 'John Cena, you're under arrest for statutory rape.'" According to employees involved in the project, Meta loosened its own guardrails to make the bots more engaging, allowing them to participate in romantic role-play, and "fantasy sex", even with underage users. Staff warned about the risks this posed. Disney, reacting to the findings, said, "We did not, and would never, authorise Meta to feature our characters in inappropriate scenarios."

Automakers are increasingly pushing consumers to accept monthly and annual fees to unlock preinstalled safety and performance features, from hands-free driving systems and heated seats to cameras that can automatically record accident situations. But the additional levels of internet connectivity this subscription model requires can increase drivers’ exposure to government surveillance and the likelihood of being caught up in police investigations. Police records recently reviewed by WIRED show US law enforcement agencies regularly trained on how to take advantage of “connected cars,” with subscription-based features drastically increasing the amount of data that can be accessed during investigations. Nearly all subscription-based car features rely on devices that come preinstalled in a vehicle, with a cellular connection necessary only to enable the automaker's recurring-revenue scheme. The ability of car companies to charge users to activate some features is effectively the only reason the car’s systems need to communicate with cell towers. Companies often hook customers into adopting the services through free trial offers, and in some cases the devices are communicating with cell towers even when users decline to subscribe. In a letter sent in April 2024 ... US senators Ron Wyden and Edward Markey ... noted that a range of automakers, from Toyota, Nissan, and Subaru, among others, are willing to disclose location data to the government.

Data that people provide to U.S. government agencies for public services such as tax filing, health care enrollment, unemployment assistance and education support is increasingly being redirected toward surveillance and law enforcement. Originally collected to facilitate health care, eligibility for services and the administration of public services, this information is now shared across government agencies and with private companies, reshaping the infrastructure of public services into a mechanism of control. Once confined to separate bureaucracies, data now flows freely through a network of interagency agreements, outsourcing contracts and commercial partnerships built up in recent decades. Key to this data repurposing are public-private partnerships. The DHS and other agencies have turned to third-party contractors and data brokers to bypass direct restrictions. These intermediaries also consolidate data from social media, utility companies, supermarkets and many other sources, enabling enforcement agencies to construct detailed digital profiles of people without explicit consent or judicial oversight. Palantir, a private data firm and prominent federal contractor, supplies investigative platforms to agencies. These platforms aggregate data from various sources – driver’s license photos, social services, financial information, educational data – and present it in centralized dashboards designed for predictive policing and algorithmic profiling. Data collected under the banner of care could be mined for evidence to justify placing someone under surveillance. And with growing dependence on private contractors, the boundaries between public governance and corporate surveillance continue to erode.

Have you heard of the idiom "You Can’t Lick a Badger Twice?" We haven't, either, because it doesn't exist — but Google's AI seemingly has. As netizens discovered this week that adding the word "meaning" to nonexistent folksy sayings is causing the AI to cook up invented explanations for them. "The idiom 'you can't lick a badger twice' means you can't trick or deceive someone a second time after they've been tricked once," Google's AI Overviews feature happily suggests. "It's a warning that if someone has already been deceived, they are unlikely to fall for the same trick again." There are countless other examples. We found, for instance, that Google's AI also claimed that the made-up expression "the bicycle eats first" is a "humorous idiom" and a "playful way of saying that one should prioritize their nutrition, particularly carbohydrates, to support their cycling efforts." The bizarre replies are the perfect distillation of one of AI's biggest flaws: rampant hallucinations. Large language model-based AIs have a long and troubled history of rattling off made-up facts and even gaslighting users into thinking they were wrong all along. And despite AI companies' extensive attempts to squash the bug, their models continue to hallucinate. Google's AI Overviews feature, which the company rolled out in May of last year, still has a strong tendency to hallucinate facts as well, making it far more of an irritating nuisance than a helpful research assistant for users.

New chat logs released by the House Judiciary Committee this week show the extraordinary lengths the FBI went to behind the scenes to shut down any discussion of Hunter Biden’s laptop in October 2020 after the New York Post broke the story. The conversations, withheld by the FBI under Director Chris Wray, show that senior leadership issued an internal “gag order” on the laptop. The FBI had been in possession of the abandoned MacBook Pro for 10 months by that stage, after computer repair shop owner John Paul Mac Isaac handed it over. The FBI’s forensic analysts quickly determined the laptop belonged to Hunter, had not been tampered with or altered in any way, and was suitable to be used in court. Yet the chat logs show that senior FBI officials instructed agents to say “No comment” when asked about the laptop during regular meetings with social media companies before the 2020 election. The FBI had spent weeks warning Facebook and Twitter about election interference in the form of Russian disinformation and had told Twitter to be on guard for a “hack and leak” operation “likely” involving Hunter Biden. In other words, the FBI “prebunked” The Post’s story so that the social media companies immediately censored it. The FBI knew The Post had received a hard-drive copy of the laptop from Donald Trump’s lawyer Rudy Giuliani because it had a covert surveillance warrant on the former mayor’s iCloud.

Since 2014, [French choreographer and dancer Alice Chauchat] has been conducting choreographic research on human togetherness amid all our differences. The result are several choreographic scores that activate paradoxical relationships: distant intimacy, attentive autonomy, impersonal engagement, pleasure and unknown play. Chauchat’s work takes place in dance studios, in exhibition spaces, on stages ... and increasingly dance gatherings in public spaces in Copenhagen, Barcelona, and of course Berlin, sprawled across the neighborhoods around where she lives. Strangers invite each other to dance as a practice of being with oneself and being with one another “without thinking about dance as self-expression, but instead thinking of dancing as relating.” I ask Chauchat what dancing means to her. “Dancing is a space where we can practice life. Because dancing is observing and being active at the same time. You perceive a situation, and you take part of it. I chose to see dancing as a form of relating. So for me it is never just dancing but always dancing with.” She sees dance as a way to offer something to someone, as a form of productive confusion, as a structure of unexamined trust. The dance gatherings, she says, give people a frame and therefore a safe space and opportunity to experience what dance can do: “You are not exposed to being a great dancer or a bad dancer. You are busy with something that relieves emotional pressure.”

Benjamin Ree’s “The Remarkable Life of Ibelin” ... is a rare and beautiful thing: a moving documentary that excavates the question of the “real” in a profoundly humanistic and unconventional way. [The film] is about Mats Steen, a Norwegian man who died in 2014 at the age of 25. Mats lived out his final years nearly immobilized, the result of being born with Duchenne muscular dystrophy. Mats left behind something for his family to find: the password to his blog. Robert and Trude, Mats’s mother, logged in to leave a note for any readers about his passing. What happened shocked them: They began to receive emails from people all over Europe, an outpouring of love and tribute to Mats, whom everyone called “Ibelin.” They were players in World of Warcraft, members of the same guild — or “community of friends,” as one participant puts it — which called itself Starlight. Mats played as a burly, friendly man he called Ibelin. Mats, as Ibelin, was involved in other players’ lives. A significant friend is Lisette, who lives in the Netherlands and met Mats in the game when they were both teenagers. Another is Xenia-Anni, a Danish mother who struggled to connect with her son Mikkel, in part because of his autism, until they met Mats in the game. Gaming wasn’t a distraction, but a life, a place where [Mats] could express all the complex parts of his personality that the physical world couldn’t accommodate. All of these people and many others tell Ree that they received wise advice from Ibelin that changed their lives. They built real friendships and had real fights and worried about one another.

A big challenge for democracies today is a decline in trust. The share of Americans who trust government, for example, has fallen from 77% to 22% since 1964. The latest attempt to build trust in the United States is a new online, state-run public forum called Engaged California. The effort aims to prompt, gather, and synthesize conversations about the state’s response to the Los Angeles wildfires into reforms. When Taiwan began a similar program in 2014, approval for the government was below 10%. Within eight years, it was 70%, although other factors contributed. The idea of designing civic spaces for civil dialogue has been best expressed in citizen assemblies. Two decades ago, for instance, British Columbia’s premier wanted to reform the electoral system but knew few people would trust the government to do it. So he recruited a wide-ranging group of citizens, asking them to devise a solution after listening to a diversity of experts. Citizen assemblies have helped build mutual trust, found Stephen Elstub, professor of democratic politics. “Because [they] require participants to listen to each other’s views and debate in an informed and reasonable way,” he wrote, “they can improve the quality of democracy.” These assemblies have been used worldwide, most notably to help Ireland navigate fraught topics such as abortion. Before they worked in such groups, 72% of participants were dissatisfied with how democracy was working. Afterward, dissatisfaction dropped to 54%.

A video game is helping thousands of vulnerable young people turn their darkest moments into moments of creativity and hope that they can pass on to others. Apart of Me is designed to help children and young adults open up about their loss and trauma in a safe and supportive environment. The game, co-founded by Manchester psychologist Louis Weinstock, is now a charity and has helped 44,000 people in the UK, and 160,000 worldwide to understand and process their grief. Created in 2018, [Apart of Me] is designed for 11-18-year-olds who have been affected by a loss which is potentially impacting their mental health. Set on an island, users can play the 3D game anonymously as it introduces characters who each have a grief-related struggle they are finding hard to deal with. It is the user's task to help the characters find a way through and this is done by collecting objects, with each object informing the user about different aspects of grief, Mr Weinstock explains. The game also allows the user to ask questions they may feel scared or uncomfortable to talk about generally. "It gives young people an outlet to have those conversations that otherwise might be difficult to have," he added. Mr Weinstock said the original idea for the game came from young people themselves as "not all want to sit in a room with a stranger and talk about their feelings". He met with young people at a hospice who all expressed the idea of a game to help them with their bereavement.