An anonymous dossier accuses PlayStation support staff of selling confidential customer data to account thieves. The corruption claims are unproven. Two years of documented takeovers do show a recovery weakness that is real, repeatable, and impossible to tell apart from an inside job when you are the victim.
How a journalist lost his PSN account twice in one afternoon
At 2:51 p.m. on December 22, 2025, Nicolas Lellouche, a technology journalist at the French publication Numerama, got an email from Sony confirming that the sign-in address on his PlayStation Network account had been changed. Eleven minutes later PayPal sent a receipt for €9.99 paid to Sony Interactive, the standard fee for changing a PSN online ID. Lellouche had run two-factor authentication on that account for years, and since Sony added passkey support in early 2024 he had signed in only with Face ID on his iPhone. It made no difference. His PS5 and his phone app were logged out, and someone using the name “Derol Bodden” had renamed the account and wiped his friends list, his conversations, his profile photo, and his country setting.
Lellouche phoned PlayStation Support, waited forty minutes, and got an agent who returned the account in under five minutes after two questions: his PSN nickname and a transaction number from an old invoice. He had it back at 5:28 p.m. At 6:04 p.m. it was stolen again, according to his own detailed account in Numerama.
The thief eventually chatted with him through PSN messaging and explained the trick without much effort. Lellouche had accidentally published a transaction number in a screenshot for an April 2023 article about Sony’s purchase-confirmation emails, and someone had saved it. A public username plus that one static number was enough to call support, pose as the owner, and have the registered email rewritten. As Lellouche summed it up on X, that “also disables 2FA.” The passkey guarded the front door, and support was authorized to move the door.
Documented cases now stretch across two years, several countries, and some of the best-known accounts on the platform. The new and unproven part is the claim that the data feeding these takeovers was bought directly from people inside PlayStation’s own support operation, and was not merely scraped from old receipts and breached mailboxes. This article examines that claim, the evidence behind it, the verified record it builds on, and the forensic work that would settle it.
The PSNINSIDE dossier: what it alleges about PlayStation support
On November 6, 2026, a now-banned X account called @PSNINSIDE, which describes itself as an independent security researcher, published a long “formal demand for an immediate internal investigation” addressed to Sony and PlayStation executive management. On November 16 it posted an update alleging “new cases of internal corruption,” with partially masked PSN IDs and screenshots that supposedly document account sales.
The central allegation is that employees with access to PlayStation’s internal customer-support tooling look up player records and sell what they find: legal names, addresses, email addresses, purchase histories, and console serial numbers, sometimes for Bitcoin. Buyers then impersonate the owner to support staff and walk away with the account. The dossier makes four further claims:
- Thousands of PSN IDs, including rare three- and four-character handles, are openly advertised on Telegram, Instagram, and WhatsApp.
- Attackers place fake callbacks posing as Teleperformance support staff or a “Senior Sony Security Engineer” to harvest verification details.
- Thieves send abusive material from stolen accounts on purpose, so the rightful owner gets banned after recovery.
- The activity spans Japan, the United States, Canada, Germany, and France.
What the dossier demands from Sony
The document asks for an independent executive investigation, preservation of access logs and call recordings, an audit of the internal tool it calls “PACMAN,” and restoration of victim accounts. It threatens public disclosure and a referral to the FBI if Sony does not act.
Why the French press urged caution
Numerama covered the dossier in mid-November, days before Grand Theft Auto VI launched, and kept its assessment sober. The accusations cannot be verified at this stage, the author is anonymous, and the author admits that the image of the internal “PACMAN” interface in the dossier is a reconstruction, not an original production screenshot.
The dossier circulates with linked archives of screenshots, marketplace conversations, and victim materials. I am not republishing those archives or the directory of marketplace handles. They purport to contain confidential customer data, and a stolen-goods bazaar does not need free advertising. What I can do is test the claims against the public record, which is substantial.
📬 Stay Ahead of Cyber Threats
Get the latest cybersecurity news, critical vulnerabilities, threat intelligence, tutorials, and exclusive giveaways delivered straight to your inbox. No spam. Unsubscribe anytime.
What Sony has said so far
Sony has not answered the dossier’s specific allegations. The nearest thing to a public response came a day before the dossier appeared. On Sony’s November 5 earnings call, CFO Lin Tao acknowledged a series of incidents, said management takes them seriously, apologized to inconvenienced customers, and declined to give specifics beyond remarking that cyberattacks are growing more sophisticated. She did not mention support staff, recovery procedures, or insider access.
Two years of documented PSN account takeovers
Hakoom: the record-holding trophy hunter who was banned after a hack
The earliest high-profile case is Hakam Karim, the Bahraini trophy hunter known as Hakoom, whom Guinness World Records lists for the most PSNProfile trophy points ever collected: 2,306,475, verified in September 2018. As Gameranx documented in a November 2024 timeline, he lost access to his account on September 9, 2024. Ten support agents would not tell him why he was locked out.
Someone then contacted him over WhatsApp and showed him evidence that his personal information was circulating online. The material claimed a “PSN admin” was compromising multiple accounts, including one belonging to a Polyphony Digital employee. While Hakoom tried to sort this out with Sony, a man calling himself “Anthony” phoned, claimed to be from Sony, and asked him to “verify” his credit card information. Hakoom recorded the calls. Sony told him it does not call customers this way.
The episode ended in a permanent ban email, not a restoration. It cited terms-of-service violations: platinum trophies earned in hours on games that normally take around a hundred, completions logged at zero minutes, more than 2,000 accounts signed in on his consoles, and sign-ins across 29 PlayStation systems. Hakoom explained the trophy timings as ordinary hobby artifacts, namely cross-generation “auto-pop” trophies and a storefront full of five-second platinum games. The login counts were never publicly judged against either his critics’ reading (the account-sharing economy around trophy hunting) or his own claim of compromise.
His account came back roughly 45 days later. On November 13, 2024, he announced his retirement from PlayStation in a post alleging corruption and sensitive-data leakage, claiming PlayStation employees had stolen and sold his personal data. In October 2025 he amplified a leaked video that purports to show the support software itself, writing that customer support has access to the “PAC-MAN system” and that “Every acc is compromised.” Notebookcheck’s reporting on the video describes an interface that shows payment information, console serial numbers, and other sensitive fields on one screen.
dav1d_123: $20,000 of games lost despite two-factor authentication
The second case put numbers on the stakes. David Tremblay, the Canadian trophy hunter known as dav1d_123, has more than 1,600 platinum trophies and had spent roughly $20,000 on PlayStation games. He lost his account on October 7, 2025, even though it was protected by Authy-based two-factor authentication. His friend, the commentator GGmuks, messaged the stolen account and got a reply from the thief, “Zzyuj.” Zzyuj claimed support could be talked into handing over an account with little more than a username, and that agents were poorly paid and poorly trained in security protocols.
PCMag covered the case in March 2026 under a headline that read like an experiment result: scammers are exploiting a security gap in the PlayStation Network to hijack accounts, and the reporter “hacked” his own account in about 30 minutes. Days later, PCMag reported an escalation. After Tremblay spoke publicly about Sony’s poor security, a hacker retaliated by duping Sony into banning his account. The theft victim became the subject of enforcement.
Nicolas Lellouche and the privileged-access clue
Lellouche’s December 2025 ordeal drew worldwide coverage because of who he was: a security-literate journalist with a passkey, a password manager, and 2FA, beaten twice in three hours by an old invoice number. Numerama’s follow-up contains a detail that deserved more attention than it got. In May 2026 a fresh attempt on his account was blocked by Sony, but an agent let slip that, at the agent’s own access level, the account had become inaccessible because it had been modified too many times. Numerama inferred that whoever made those modifications was working with privileges above those of a standard support agent. That is an inference, and a reasonable one.
Colin Moriarty and RealRadec in 2026
On May 18, Colin Moriarty, host of the Sacred Symbols podcast, former IGN editor, and Kinda Funny co-founder, posted that his account had been hacked despite 2FA. Another hacked user had warned him shortly before that “They have your information and they are going to try to take your account today.” The compromised account was immediately used to send his co-host Dustin Furman a PSN message that read “You’re next.”
Support first told Moriarty restoration could take three weeks. He escalated through PlayStation PR contacts and friends at Sony-owned studios and had the account back the same day. He acknowledged afterward that he got that treatment only because of his standing in the PlayStation community, a privilege most of the platform’s 127 million monthly active users do not have. Moriarty also speculated publicly that “there is some sort of insider access behind this.” That is a suspicion, not a finding. Kotaku noted that Sony did not respond to its request for comment.
In early October 2026 the PS5 creator RealRadec was hijacked and later regained access in what was described as a near-identical case.
What independent reporting adds
The well-known PlayStation commentator @mrpyo1 summarized the attack in May 2026. Attackers need only your public PSN ID plus one piece of old transaction data, usually a full order number or the last four digits of any card ever used on the account. Support accepts that as “proof of ownership” before changing the registered email, which disables 2FA.
A June 2026 report by the Polish magazine CD-Action drew on interviews with victims and with Ginko, the world’s number-two trophy hunter. It describes an organized crew mass-attacking top trophy hunters, naming Hakoom, Matrixlord, dav1d_123, and TJMEAT, the leading American trophy hunter, among the victims. It also shows an account thief known as “Nich” boasting in a video reel about stealing a top hunter’s account.
Ginko told CD-Action that 2FA and passkeys are being bypassed wholesale. Polish victims who phone support tend to recover quickly, he said, while North American victims are often told to create a new account. Hackers also file theft reports on accounts they have seized, which poisons the recovery process against the true owner.
Matrixlord, whom the dossier names as a victim of data disclosure and sale, shows up in independent coverage as a target of the same campaign. The specific claim that an insider sold his records still rests on the dossier alone.
As of May 2026, wccftech reported that the underlying support weakness remained unaddressed six months after Lellouche’s story. In November, Numerama stated that, to its knowledge, Sony still had not changed the procedure. One report suggests Sony may be working on the problem quietly: Windows Report, citing PushSquare, noted two PlayStation job listings for a Senior Fraud Strategist, in California and London, after the May reports. The listings do not mention these incidents, and nothing confirms a link.
The timing matters. Grand Theft Auto VI launched on November 19, 2026. The UK Information Commissioner’s Office issued a security warning to gamers on November 16, and researchers documented a 1,300 percent surge in gaming-themed phishing ahead of launch, with GTA impersonations making up 42.4 percent of phishing emails in one analyzed weekend. The account-recovery desk is now serving the largest launch audience in the platform’s history.
How an insider-assisted PlayStation account takeover would work
The dossier’s most technically consequential claim concerns structure, and the structure matters more than the claim that insiders exist. The takeover is split into two transactions that never touch each other in the logs.
First, an employee with legitimate read access to support tooling queries a target’s record and extracts exactly what the verification script asks for: the registered email, date of birth, a recent order number, the last four digits of a linked card, and the serial number of the console that created the account. Nothing is modified. The employee sells this packet, typically through encrypted messaging and for cryptocurrency.
Second, the buyer contacts PlayStation Support through an ordinary channel, impersonates the owner, and answers every ownership question correctly, because the answers came from the system of record. On the dossier’s account, a second and unwitting agent processes a routine recovery: the registered email changes, credentials reset, control transfers. The case closes as a legitimate support interaction.
The elegance here is forensic. An investigation that begins at the recovery transaction will find an agent who followed the script against data that matched. The insider who made the original read appears nowhere in the recovery case, since reading a record and modifying it are different events, often done by different people days apart. That is why the dossier insists, correctly as a matter of method, that examining the recovery agent alone is not enough. The first question has to be who searched, viewed, copied, or exported the record before the recovery attempt, under which employee ID, with which permissions, and whether the access had any operational justification.
PACMAN: the PlayStation support tool Sony has never confirmed
The name comes from the community. Hakoom’s October 2025 post calls it the “PAC-MAN system,” Notebookcheck describes a leaked video of support software displaying payment data and console serials, and the dossier brands its recreated interface image with the same name. Sony has never publicly confirmed the tool’s name, architecture, or permission model, and the dossier’s author concedes the image is a reconstruction.
The capability class is beyond dispute even if the branding is not. Lellouche’s hacker told him plainly that customer service has a tool that resets the registered email even when the account is protected by a password or a passkey. Every major consumer platform runs such a console. Twitter’s 2020 attackers reached internal account-management tools, Microsoft’s 2019 incident involved a support agent’s credentials, and Coinbase’s 2025 breach involved support agents’ view of customer records. The documented takeovers establish that functionally equivalent power exists at Sony. What remains open is who can use it, on whose records, with what logging, and whether anyone has ever audited the reads.
Why a passkey cannot protect a PSN account
The failure is architectural, and it predates any allegation of corruption. Authentication and recovery are separate systems, and Sony’s investment in the first (passkeys since February 2024, two-step verification before that) does nothing for the second. Sign-in asks what you possess, a cryptographic key bound to your device. Recovery, as practiced against these victims, asks what you know: a username that is public by design, an order number printed on every receipt email Sony sends, card digits that surface in any financial breach, and a console serial photographed on resale listings and printed on the box. These are static secrets that cannot be revoked.
Lellouche asked support whether the compromised transaction number could be invalidated so the hacker could not reuse it. He was told no such mechanism exists. The secret that stole his account in December 2025 was minted in 2020 and stays valid indefinitely.
Standards bodies have spent a decade trying to retire this model. NIST’s digital identity guidelines, SP 800-63B, say in normative language that verifiers “SHALL NOT prompt subscribers to use knowledge-based authentication.” Security questions and their relatives have answers that can be guessed, obtained, or, as here, read by the same organization asking the question. An insider with record access turns knowledge-based verification into a tautology: the attacker proves ownership by reciting the contents of the owner’s file, and the file is what the insider sold.
Even without an insider, the documented cases show a verification floor low enough that one leaked artifact is enough, with basic velocity checks missing. Numerama’s journalist pointed out that support saw no anomaly in three different people claiming the same account within three hours.
Insider threats at the support desk: Twitter, Microsoft, MGM, and Coinbase
The threat class is real outside gaming too, and it concentrates at the support desk.
In July 2020, attackers called Twitter employees pretending to be the company’s IT department and talked four of them out of their credentials. They used internal tools to hijack accounts belonging to Barack Obama, Elon Musk, and cryptocurrency exchanges in a Bitcoin doubling scam that took in more than $118,000. The New York Department of Financial Services’ investigation found Twitter lacked adequate access controls, identity management, and security monitoring, and had no CISO at the time.
In April 2019, Microsoft disclosed that a support agent’s credentials had been compromised. Outsiders could read customer email account metadata and some message content for roughly three months. The breach hit the support layer and left the mail platform itself untouched.
In September 2023, MGM Resorts lost an estimated $100 million to attackers who, as CISA’s Scattered Spider advisory describes, convinced IT helpdesk staff to reset passwords and MFA tokens. The helpdesk was the perimeter that failed.
The Coinbase insider case
The closest analog to the dossier’s allegations is Coinbase. In a May 14, 2025 securities filing, the exchange disclosed that criminals had bribed overseas support agents to exfiltrate customer records (names, addresses, masked identifiers, account data) starting in December 2024, affecting nearly 70,000 customers, for use in impersonation attacks. The attackers demanded $20 million. Coinbase refused, posted a $20 million reward for information leading to convictions, and put remediation costs at $180 million to $400 million.
A September 2025 class-action filing in the Southern District of New York, as reported by Infosecurity Magazine, identified the agents as employees of the outsourcing firm TaskUs in India. It alleged that one employee, Ashita Mishra, photographed up to 200 customer records a day and sold them for $200 each, amassing over 10,000 before her January 2025 arrest, and that she recruited supervisors and team leaders into the scheme. TaskUs confirmed staff involvement and disputed the scope.
Every structural element in the PSN dossier (privileged reads, monetized disclosure, downstream impersonation, an outsourced workforce, supervisory blind spots) exists as proven fact in the Coinbase case. That does not make the Sony allegations true. It makes them plausible, and it sets the standard of investigation they deserve.
The black market for rare PSN IDs
Scarcity and status drive this ecosystem. PSN online IDs run from three to sixteen characters, and the shortest handles, the three- and four-letter “OG” names claimed in the network’s earliest years, are finite and unrenewable. In the community they work like jewelry. Sony’s terms of service forbid sharing, buying, selling, renting, trading, or transferring accounts, account details, or credentials in plain language.
When Sony allowed online ID changes in April 2019, with the first change free and later ones $9.99 ($4.99 for PlayStation Plus members), it handed account thieves a price list and a tell. The €9.99 that alerted Lellouche was his thief renaming the account he had just stolen.
The dossier describes a semi-public bazaar: Telegram channels with names like PSN Market and PSN Open Market, Instagram storefronts advertising rare handles, WhatsApp listings shown in Numerama’s coverage, and settlement in Bitcoin. That matches what independent reporters have seen. CD-Action’s report shows “Nich” bragging in a reel about a stolen trophy-hunter account, and Lellouche’s December article referenced dedicated Telegram groups for account theft.
Two cautions apply. Sellers’ claims of insider access are marketing as much as confession, and the OG-handle underworld has always inflated its own mystique. Lellouche’s hacker needed nothing more than a published invoice number and no mole at all. From the victim’s side, a pure social-engineering takeover and an insider-assisted one produce identical symptoms, so only Sony’s internal access logs can tell them apart.
The payment rail is also less anonymous than sellers assume. Bitcoin is pseudonymous, and chain analysis followed by exchange subpoenas has unwound far larger operations. Telegram changed its privacy policy in September 2024 to disclose user IP addresses and phone numbers to authorities on valid legal requests in criminal cases, and it reportedly shared data on 2,253 users with U.S. authorities in 2024. A coordinated investigation has more to work with than the sellers believe.
Outsourced support, fake callbacks, and Teleperformance
The dossier names Teleperformance, the French customer-experience outsourcing giant, in a specific way: attackers allegedly impersonate its support personnel on callbacks, alongside the “Senior Sony Security Engineer” persona. Sony does not publicly disclose its support vendors, so an impersonation claim does not prove Teleperformance staffs any PlayStation program. Teleperformance is, however, a major provider of outsourced player support to the gaming industry, and the distinction matters little in practice because a victim cannot tell a real vendor callback from a fake one.
Hakoom’s recorded “Anthony” calls, asking him to verify credit card details, are the documented version of this vector, and Sony itself told him it does not make such calls. Numerama’s advice to readers was blunt: if a “Sony security engineer” phones unprompted and asks for your console serial number, hang up. You may be talking to the person who bought your account.
The outsourcing layer still deserves scrutiny on its own record, stated carefully. Teleperformance USA reported a data breach to the Texas Attorney General on December 8, 2022, exposing names, addresses, and Social Security numbers, with 2,631 affected individuals in Texas alone. Nothing public connects that incident to PlayStation data, and it should not be presented as if something did.
Teleperformance’s own 2023 fraud white paper lists as a canonical insider scenario “an employee collects social security numbers and resells them to an accomplice.” The industry knows what this attack looks like. The TaskUs case shows what happens when that knowledge does not become field-level access controls, read logging, and behavioral monitoring. The governance instruments to force the conversion already exist: GDPR Article 28 processor obligations, PCI DSS service-provider management requirements, ISO 27001 screening and logging controls, and contractual audit rights. Only Sony and its vendors can say whether their contracts for the recovery desks enforce them, and the question falls squarely within the investigation the dossier demands.
When hackers get victims banned: moderation as a weapon
The dossier’s cruelest allegation is that thieves weaponize enforcement. After seizing an account, they send abusive or obscene material from it to plant a violation record, so the legitimate owner is suspended for content the attacker created. That precise form of the claim is unverified, but adjacent behavior is documented:
- Moriarty’s stolen account was used to threaten his co-host within minutes of the takeover.
- Tremblay’s public criticism of Sony’s security was answered, per PCMag’s March 2026 reporting, with a hacker duping Sony into banning his account.
- Ginko told CD-Action that hijackers file theft reports on the accounts they steal, turning the recovery system against the owner.
- Lellouche’s account was suspended mid-tug-of-war with a “red code” on his profile, while he waited five to ten days for a resolution.
The pattern is consistent. Sony’s enforcement and recovery machinery treats the account as the actor and cannot currently distinguish the owner from the occupant.
The fix is forensic. Every enforcement action should be correlated against authentication telemetry (login timestamps, device identifiers, IP geolocation, session origins) and against any acknowledged compromise window. Content sent from a session the owner demonstrably did not control should not produce standing consequences for the owner, and suspensions issued without that correlation should be re-reviewed once a compromise is established. This is ordinary log analysis on data the platform already holds. On the public evidence, what is missing is a process that uses it.
What a credible Sony investigation would look like
The dossier’s fifteen demands reduce technically to a coherent program.
Preserve the logs and trace every read
It starts with preservation: an immediate litigation hold over support-tool access logs, customer-search histories, record-view and export events, case notes, recovery and verification records, email and ID change histories, call recordings, supervisor approvals, and authentication and device telemetry for every affected account. The window before each takeover gets priority, before routine log rotation erases the only evidence that matters.
The analytic core is full-chain attribution. For each compromised account, investigators reconstruct who read the record before the recovery call, under which employee ID and permission set, whether the read had a legitimate case justification, and whether the same reader recurs across unrelated victims. A read with no operational reason, hours before a stranger recovers the account, is the signature the dossier alleges. That signature either exists in Sony’s logs or it does not, which makes this investigation more tractable than most breach attribution.
Fix the recovery architecture
The preventive design is well understood:
- Least-privilege roles, so a billing agent cannot see console serials and a troubleshooting agent cannot see full payment history. Sensitive fields stay masked by default and unmask only with a logged justification.
- Just-in-time elevation, a second approver, and a cooling-off period for credential resets and email reassignments, with notification to the previously registered address and hard velocity limits on how often one account can be “recovered” in a day.
- Verification based on possession and proofing instead of static knowledge: device-bound confirmation where possible, document-based identity proofing for high-value or contested accounts, and a way to invalidate leaked artifacts, the absence of which Numerama’s reporting exposed.
- Behavioral analytics that flag anomalous read patterns, plus canary records (synthetic accounts with no legitimate reason to be opened) that turn insider browsing into a tripwire.
- An in-app way to authenticate inbound callbacks, which would kill the fake-engineer vector.
- The compromise-window correlation for moderation described above.
Sony’s legal exposure
The law Sony operates under makes this mandatory. If staff accessed and disclosed personal data, GDPR Articles 33 and 34 require notice to supervisory authorities within 72 hours of awareness and to affected users where risk is high, with fines up to €20 million or four percent of global turnover. Japan’s amended APPI has required breach reporting to the Personal Information Protection Commission since April 2022. California’s CCPA gives breached consumers statutory damages of $100 to $750 each where reasonable security was lacking.
Sony has been here before. The 2011 PSN intrusion exposed personal data on 77 million accounts and took the network down for 23 days, and the UK Information Commissioner fined Sony £250,000 in January 2013 over it. A substantiated insider-disclosure scandal would be worse in kind, because it would betray the recovery function users are told to trust. The dossier’s threat to go to the FBI and data-protection authorities describes where such matters end up when companies sit on them.
For the result to carry weight, the investigation has to be structurally independent of the support organization it examines, cover every region the dossier names, and end in a published account of findings. Anything less will read, fairly, as the department investigating itself.
What is proven and what is still alleged about PlayStation insider selling
Several things are established by multiple independent outlets and the victims’ own records:
- PlayStation’s support recovery process has, for at least two years, handed account control to impostors holding a public username and one static artifact, defeating passkeys and 2FA.
- The artifacts cannot be revoked.
- Velocity checks failed even on a journalist’s account claimed three times in a single afternoon.
- Enforcement has punished victims, including at least one retaliatory ban.
- Sony acknowledged “a series of incidents” in general terms on its November 5 earnings call. It did not address the support process or any specific case, did not answer Kotaku’s request for comment, and had not, per Numerama, changed the procedure as of mid-November 2026.
What is alleged, by an anonymous dossier whose centerpiece image is an admitted reconstruction, is that corrupt insiders read and sell the records that make these takeovers possible, at scale, across five or more countries, for Bitcoin. Neither Sony nor Teleperformance has responded to those specific claims in any reporting I reviewed. Both were entitled to, and this publication would carry their answers.
The two hypotheses, a support system lax enough to be socially engineered and one actively leaking from within, look identical from where victims stand, and Sony alone holds the evidence that separates them. Every recovery described here generated logs showing who read the record, who changed it, from which workstation, and who approved it. If the dossier is wrong, those logs exonerate Sony, and publishing the audit is the cheapest reputational repair available. If it is right, each day the logs go unexamined is another day the keys are for sale. A platform that has just taken on the launch audience of the most anticipated game in its history, and that asks users to keep their entire digital libraries behind one account, cannot leave the question open indefinitely. Sony holds the access logs for every recovery case above, and an audit of who read each record beforehand would settle it.








