PWNED Welcome back to PWNED, the weekly column where we gather lessons from organizations that didn’t take security seriously enough. This week’s tale of woe comes from a company that left a door wide open for miscreants, but was lucky it didn't have to pay the price. Have a story about someone leaving a gaping hole in their network? Share it with us at pwned@sitpub.com. Anonymity is available upon request. Our story comes courtesy of a reader we’ll Regomize as Manny. A few years ago, Manny got a job working at a law firm. The firm used him to replace an entire team, making him the de facto IT department all by himself. He soon discovered that all of the company’s data and applications lived in one large web-based interface, which was divided up based on the type of client. So there were areas in the UI for personal injury cases and others for travel refunds, for example. There was just one big, gaping security hole: a master password that allowed you to log in as any user in the system. If you had this password, which many people in the law firm did, you could grab detailed personal information about any client, even their health records. “I immediately raised this as a huge security risk,” Manny told us. “But I was told, 'Oh that's the admin password, everyone uses it. Don't touch it.'” As long as you had the person’s email address that you wanted to impersonate, this password would allow you to impersonate them. This applied to both staff and clients. “Colleague is off sick? Sign in as them and reassign their work to someone else to complete. Client forgot to fill in a field? Log in as them and complete it for them,” Manny said. The system itself was 15 years old, ancient in tech terms, and it desperately needed replacing. So Manny was asked to build a whole new system. Naturally, he refused to add a back door, even though that’s what the boss wanted. “I point blank refused to add any back doors to it,” Manny recalled. “So they promoted every user to a system admin and carried on, business as usual.” What we can take away from Manny’s experience is that sometimes even the best IT people who know security basics can still be hindered by clueless management. We also know that sometimes in order to pay the bills, IT people have to go along with security practices they strongly disagree with. In the end, the boss will have the final word, even if that word is “ignorance.” ®
Australia’s Privacy Commissioner has revealed a tech support scam was the cause of the massive 2025 data breach at Australian airline Qantas and found the carrier didn’t breach its privacy obligations despite leaking personally identifiable information for 5.7 million customers. The Commissioner reached that conclusion, and a decision not to open a formal privacy probe, in a report published today. Qantas has previously admitted the incident was the result of a social engineering attack on a contact center. The Commissioner’s report goes deeper, explaining a crook who claimed to represent “Qantas IT help” made the call and told a contact center agent to access a CRM system and perform certain actions needed to close a support ticket. Those actions instead connected the CRM to a data extraction tool which the crooks used to siphon off customer records. The Commissioner considered whether Qantas observed the Australian Privacy Principles (APPs), the binding rules that govern how businesses safeguard PII, and found the airline did the right thing. The report found that Qantas audited the operator of the contact center and tested the security awareness of its employees – and had done so in the months before the incident. Qantas also conducted mandatory and recurring training on how to handle PII. The Commissioner was therefore satisfied Qantas took adequate steps to ensure the contact center observed the APPs and didn’t fail in its obligations. The regulator made a similar finding regarding the airline’s cross-border data-sharing practices. “Our inquiries did not identify any omissions in the steps Qantas took that, if addressed, would have prevented the breach that occurred in this incident,” the report states. The APPs include a requirement to take reasonable steps to protect personal information from unauthorized access. Again, the Commissioner decided Qantas complied because it used role-based access controls, among other techniques to protect data. Another issue the regulator considered was whether Qantas took reasonable steps to destroy or de-identify the personal information it didn’t need. The carrier told the Privacy Commissioner that it scheduled annual data removal runs from its CRM, and that no records that deserved deletion or removal were present at the time of the attack. That clean record saw the Commissioner decide not to launch a deeper investigation. “I have a broad discretion to commence an investigation of an act or practice where it may be a contravention of the APPs and where it is desirable to do so,” the report states. The first-person pronoun is presumably the work of Commissioner Carly Kind, who observed “it does not appear that Qantas could have reasonably foreseen and prevented the breach in the manner that it occurred. The way in which the threat actor gained access was through a vishing attack which could not have been prevented by a strengthening of Qantas’ current role-based access controls.” It’s possible the Commissioner will revisit the matter at another time, and class-action lawsuits are also in train regarding the incident. Qantas may therefore still have to fight through plenty of turbulence before this matter lands. One thing the report doesn’t address is the identity of the attackers. Pundits have suggested the Scattered Spider gang did the deed after it started attacking the aviation industry in the weeks before the Qantas incident. ®
The crippling high-consequence attack on vital infrastructure that cybersecurity experts have warned about for years is upon us, in the form of an incident that may force KFC to close some stores in Japan. Colonel Sanders himself is not the victim here. That role goes to Nichirei Group, a Japanese purveyor of frozen foods and super-chill logistics services that move them around. Nichirei Group on Monday posted a notice in which it admitted “system failures caused by unauthorized access have occurred.” The failures meant the frozen food concern could not arrange shipments to or from its refrigerated warehouses or conduct its other operations. Shortly after Nichirei Group revealed its difficulties, KFC Japan warned customers that delivery of ingredients to its stores would likely be affected. The chicken chain therefore stopped taking orders through its app and website and said it may need to limit menu items and opening hours. “Some stores may be closed depending on the availability of ingredients,” the company said. On Wednesday, Nichirei Group confirmed the cause of the outage was a cyberattack and admitted attackers accessed a server that stores personal information. The Group declined to offer any detail on the incident “to prevent further damage.” The company hopes to resume operations on Friday. That Nichirei Group is unable to provide some services suggests a ransomware attack has made some data unavailable. The mention of “further damage” suggests that discussing whatever happened could divulge clues about security weaknesses that would allow further attacks, perhaps directed at the Group’s clients. KFC Japan hasn’t posted any information about store closures. Indeed, the company continues to promote summer menu items such as a Japanese-style citrus and chicken combo that the chain says is refreshing to eat even in the heat of summer. The Register’s Asia-Pacific bureau will not venture to Japan to assess the impact of this incident, or try the burgers: At times like this, with critical infrastructure under stress, that’s just the right thing (not) to do. ®
Since WIRED reported on Meta’s NameTag face recognition system, company executives have made confusing and conflicting remarks about its very existence.
The US Cybersecurity and Infrastructure Security Agency (CISA) has urged all organizations running SharePoint to harden their defenses after the disclosure of actively exploited vulnerabilities. The warning applies to those running any supported version of SharePoint Server on-prem, with three vulnerabilities of particular interest cited. A spoofing bug, CVE-2026-32201 (6.5), was the first to be mentioned. Microsoft disclosed it in March and CISA confirmed it was being actively exploited in June. Additionally, CISA appears concerned by CVE-2026-45659 (8.8) – a remote code execution (RCE) flaw made public in June and confirmed as being actively used in attacks last week after Microsoft said exploitation was "less likely." The most recent of the three, CVE-2026-56164 (5.3), a privilege escalation flaw, was one of the 622 bugs that featured in this month's record Patch Tuesday. CISA also picked out two more critical bugs, both from the latest Patch Tuesday, as ones that could potentially complicate SharePoint security further. Neither CVE-2026-55040 (9.1) nor CVE-2026-58644 (9.8) is being actively exploited to date, although Microsoft has attached the "Exploitation More Likely" label to both. CISA said the three exploited vulnerabilities are associated with post-exploitation activity, including the theft of Internet Information Services (IIS) machine keys and deserialization techniques, both in an effort to gain persistence and deploy malware. The agency did not offer any more detail about what led it to issue the warning, but went on to encourage defenders to review an alert it published in August 2025, which similarly urged organizations to harden SharePoint from "ToolShell" attacks. CISA said attackers were chaining together CVE-2025-49706 (6.5) and CVE-2025-49704 (8.8) to break into SharePoint Servers and, in some cases, deploy Warlock ransomware. It did not go as far as attributing the activity referenced in either SharePoint advisory to any group or country, although Microsoft said as far back as July 2025 that ToolShell vulnerabilities were being exploited by Chinese nation-state crews. Applying Microsoft's latest security patches and verifying that Antimalware Scan Interface (AMSI) integration is enabled for each SharePoint web application are among the recommended hardening measures. CISA also advised defenders to go threat hunting for signs of intrusion before rotating IIS keys to avoid exposing SharePoint to the web unless it's necessary and block external access to SharePoint Central Administration. As is the case with any potential intrusion, CISA encouraged organizations to implement robust, tailored logging that can detect potential exploits. ®
UPDATED Patch Tuesday was followed by Oopsie Wednesday for some Dell customers, with Microsoft slamming on the update brakes after the hardware maker reported some problems. Yesterday was Microsoft's monthly security update for Windows. This month was, by all accounts, a bit of a doozy with a record-breaking number of CVEs patched, some of which were classed as critical and under active exploitation. Better get patching then? Well, er, no. Not if you're using a Dell device affected by issues associated with the update. Microsoft admitted it affected "some Dell devices with Intel processors," but stopped short of providing a full list. The Register asked the Windows giant and Dell which models had been hit, but both have yet to respond. Microsoft confirmed on its update page: "This update might not be available for a limited number of Dell devices with Intel processors due to an incompatibility reported by Dell that can potentially cause unexpected shutdowns, poor performance, increased heat, and battery drain." And the fix? "We are working together with Dell to prevent the affected models from experiencing the issue and plan to release a resolution for affected devices in the coming days." While the pair works on a solution, the update is "temporarily unavailable." Thanks to the sheer number of CVEs in the update, the delay is unfortunate, doubly so when considering that only a week ago, Microsoft was fiercely advocating for users to get patches installed as soon as possible due to the speed at which AI systems can detect and exploit vulnerabilities. In this instance, Microsoft has acted quickly to halt the update for affected devices. However, the fact that it got this far and can cause surprise shutdowns, overheating, and performance problems does not speak well of the company's validation and quality procedures. Dell is hardly a bit player in the hardware ecosystem. Somewhere, deep in the heart of Microsoft's Redmond campus, a sad-faced engineer is likely resetting the "Days since we broke something" counter and thinking fondly of the days when the number reached double or triple figures. ® Updated to add on July 16: A Microsoft spokesperson told The Register: "We are aware of the issue and have paused the Windows 11, versions 25H2 and 24H2 (KB5101650) update for impacted devices while we work on a resolution." The company has also updated its Windows release health dashboard. According to the Windows behemoth, the issue occurs on Dell models with an Intel Innovation Platform Framework Processor Participant driver and is related to the new Windows USB-C Connection Manager interface. The problem first cropped up in the preview update on June 23, ahead of patch Tuesday. There's no workaround, and Microsoft has halted the rollout to affected devices until a fix is ready, "in the next few days."
UPDATED: Microsoft’s worst nightmare - a prolific zero-day vulnerability hunter who calls themselves Nightmare Eclipse - published yet another zero-day on Tuesday, a vulnerability allowing attackers to mount user hives, including partial exploit code. Suspected of being a disgruntled former Microsoft engineer, based on the sophistication of their prior vulnerabilities, NightmareEclipse came good on their promise to release another zero-day on July 14. Whether it lives up to the promised “bone-shattering” standard touted in June is up for debate, however. Called “LegacyHive,” the proof of concept (PoC) code for the zero-day local privilege escalation (LPE) vulnerability targets Windows’ user hives - the section of the Windows Registry that stores a user's specific desktop settings, application preferences, and environment configurations. The code exploits a weakness in profsvc, the Windows User Profile Service, and the way in which it loads hives. If exploited correctly it could grant regular users privileged read-write access to target other users' hives. Matei Badanoiu, lead security researcher at Pentest-Tools.com, said that while the exploit could prove useful for attackers who had already gained a foothold in a target environment, it falls short of providing a fuller system compromise. “What caught my attention is the difference between what the public proof of concept actually demonstrates and what a full compromise would require,” he told The Register. “LegacyHive is a local privilege escalation in the Windows User Profile Service. It abuses arbitrary registry hive loading, so a standard user can mount another user’s hive, including an administrator’s, into their own classes root. “For an attacker who already has a foothold, that is a genuinely useful primitive. Bundling it with credential access and persistence into ‘full compromise’ is more of an ambition than the released code.” The LegacyHive publication differs from some of NightmareEclipse’s earlier drops in that the PoC code is stripped back in an effort to prevent widespread exploitation. According to the bug hunter, there is more than one way of exploiting the profsvc flaw. The public PoC requires additional user credentials for it to work, and is limited to the usrclass.dat hive. NightmareEclipse said the original PoC, which differs from the one they published, does not require additional user credentials to exploit the bug, and it works beyond the usrclass.dat hive, “but you would need some brain cells to make the PoC do it.” This represents a divergence from NightmareEclipse’s previous approaches. As Badanoiu pointed out to us, some of NightmareEclipse’s earlier drops, such as BlueHammer and RedSun, went from PoC to widespread exploitation within days. LegacyHive, however, comes without a fully working PoC and a CVE identifier. Regardless, security experts told The Register that cyber practitioners should respond promptly since capable attackers could probably build a reliable exploit, despite the gaps left in the PoC by NightmareEclipse. “Threat intelligence teams are advised to act with some urgency here,” said Dray Agha, senior manager of security operations at Huntress. “Huntress observed NightmareEclipse's prior LPE and defence evasion tools rapidly deployed threat actors and ransomware groups shortly after publication. “Given this history, we’d expect that capable actors will reverse-engineer the missing components of the LegacyHive PoC to build fully weaponized versions in short order.” The timing NightmareEclipse may have changed their approach to releasing full working PoCs to the public, perhaps a reflection of Microsoft’s suggestion of preparing legal action against the bug hunter, but the nuisance timing of the vulnerability disclosures remains. They dropped the details for LegacyHive shortly after Microsoft released its monthly Patch Tuesday updates, which contained an unprecedented 622 fixes. Agha said timing the disclosure in this way maximizes the exposure window before a patch can be developed, causing more trouble for Microsoft. The Register asked the Windows-maker about LegacyHive and whether it was planning to release a fix before August’s patches. NightmareEclipse claims their latest zero-day works against Windows machines that are fully patched according to July’s fixes. Microsoft previously issued a quiet remedy for one of NightmareEclipse’s earlier zero-days, RoguePlanet, last week, although the company did not go into any details about what the mitigation entailed. ® Updated to add on July 16: A Microsoft spokesperson got in touch with The Reg to say is is: "aware of the reported vulnerability and is actively investigating the validity and potential applicability of these claims." They added: "Microsoft is committed to investigating security issues and updating impacted products to protect customers as soon as possible. Importantly, we support coordinated vulnerability disclosure, an industry standard that protects customers and supports the research community by ensuring their findings are thoroughly investigated and addressed before being made public."
ESET researchers discovered 11 vulnerable UEFI shim bootloaders signed by Microsoft that allow attackers to bypass UEFI Secure Boot by exploiting decade-old vulnerabilities
"Build a smart home", they said. "It'll make life so much better", they said. Well, life wasn't very bloody good at 23:00 the other night after travelling 33 hours from Paris only to find the IoT doorlock batteries dead and the 9V "jump start" procedure completely failing! Eventually, the locksmith arrived and opened an old-school physical lock on another door in an alarmingly short time. So, lessons:
Battery-powered locks suck and will eventually lock you out of your house
Don't trust a fallback mechanism as rudimentary as "hold a 9V battery on some terminals"
Always have an old school manual backup approach, AKA "a key"
As I say in the video, we do have other doors that have keys, and if it weren't for the complacency we developed, we would have had one of these accessible. But alas, we didn't. The path forward is to take a deep dive into Ubiquiti's Access ecosystem, which I've flagged in the past, and by pure coincidence, I already had a meeting lined up with them to discuss just this. So, the hardware is on the way, and I'll have something entirely new to play with in the coming weeks. Stay tuned!
Remember last month when we were awed by Microsoft’s record-setting Patch Tuesday that addressed 206 CVEs? That was a quaint era compared to this month: Redmond just rolled out patches for 622 CVEs specific to its products, slightly more than tripling last month’s all-time high. Redmond’s Patch Tuesday release is once again one for the record books, with everything under the sun getting some security fixes – including 428 non-Microsoft Chromium CVEs affecting Edge that aren’t included in that 622 count. Fifty-eight of those are critical, two are under active exploit, and one has already been publicly disclosed, meaning it could join those other two in short order. There is a lot to dig through, and we can hardly cover the whole gamut given the size of this release. As we noted last month, there was concern in the infosec community that AI-enabled bug hunting might mean massive patch volumes are the new normal. Microsoft didn’t disclose how much AI may have contributed to the massive patch list this month, but given the volume it’s safe to say human contributors probably had some assistance. Microsoft’s massive month To start, let’s cover the pair of actively exploited issues that Microsoft patched. The first, CVE-2026-56155, is an Active Directory Federation Services elevation of privilege vulnerability. Attackers who exploit the issue, which Microsoft only described as being due to “insufficient granularity of access control on ADFS,” could gain administrator privileges. They do need to have access already and be local, however, which is why this is only rated with a CVSS score of 7.8. The second actively exploited vulnerability, CVE-2026-56164, is another privilege elevation issue, this time in Microsoft SharePoint. SharePoint is apparently missing authentication for a critical function, which could let an unauthorized attacker on a network elevate their SharePoint permissions. As with the other issue under exploit, this one is somewhat limited, earning it a CVSS of just 5.3. With both under active exploitation, that score doesn’t matter as much as eliminating the vulnerability through good patch management, however. As for the publicly reported but not-yet-exploited issue, CVE-2026-50661, that involves BitLocker being able to have its security measures physically bypassed by anyone with local access to a BitLocker-secured machine. Now let’s round up a few of those 58 critical issues. Everyone’s favorite untrustworthy AI is packing a CVSS 9.6 remote code execution vulnerability. CVE-2026-48561 finds Copilot improperly neutralizing its input, allowing an unauthorized attacker to execute code with nothing but low-privileged Hyper-V guest access. Exploiting the vulnerability can be done without user awareness by, for example, hosting a malicious website that prompts the many embedded Copilot features of Windows machines to process a prompt upon landing on the page. Microsoft Exchange is suffering from a CVSS 9.6 spoofing vuln due to failure to neutralize input, leading to cross-site scripting being possible from within a maliciously crafted email. CVE-2026-55008 allows an unauthorized attacker to perform spoofing over a network by sending said malicious email to a target, allowing arbitrary JavaScript execution. Finally, we’re not picking out one vulnerability for your third notice in this massive list, but are highlighting a full 16 remote code execution vulnerabilities in Microsoft Office and its associated applications. They’re caused by a variety of issues in the Office suite, like heap-based buffer overflow and use after free vulns, and all are scored around a CVSS 7.8. Needless to say, we recommend following Microsoft’s advice and getting all those hundreds of security patches installed ASAP. Adobe throws mud at critical issues in multiple products Microsoft tends to command the headlines on Patch Tuesday (it’s hard not to when you address more than 600 CVEs in a single day), but Adobe released a bunch of patches across its ecosystem too, 64 unique CVEs across seven bulletins for Commerce, Experience Manager, Creative Cloud Desktop, Illustrator, Content Credentials SDK, ColdFusion, and Animate. Every one of the bulletins included at least a couple of critical CVEs. The highest-severity issue among Adobe’s many Patch Tuesday entries comes in the form of a CVSS 9.9 path traversal vulnerability in ColdFusion that can allow arbitrary code execution. CVE-2026-48318 does not yet appear in online CVE directories, but even with limited information, we’d say a 9.9-level issue is one you want to address with a quickness. The second-worst issue that Adobe addressed today is in its Commerce suite. CVE-2026-48356 is a CVSS 9.6 privilege escalation vulnerability that an attacker can trigger thanks to Commerce failing to restrict the upload of dangerous file types. Adobe Experience Manager also includes a pair of CVSS 9.6 issues (CVE-2026-48259 and CVE-2026-48359). Both allow arbitrary code execution: one because of a server-side request forgery vulnerability, and the other because of improper restriction of XML external entity references. Other notable Patch Tuesday releases Broadcom addressed seven CVEs in its Avi Load Balancer today, which it rates from 7.1 to 9.8 on the CVSS scale. The vulnerabilities include authentication bypass, RCE, privilege escalation, and directory traversal. SAP published 16 security updates and one GitHub advisory today; nine of those updates have a CVSS score of 8.1 or higher. CVE-2026-44747 (CVSS 9.9) is a memory corruption issue in SAP NetWeaver Application Server ABAP that could allow an authenticated attacker to gain unauthorized access to system data; CVE-2026-27690, CVSS 9.1, would let an unauthenticated attacker smuggle an HTTP request through SAP Approuter leading to system unavailability; and CVE-2026-44761, CVSS 9.1, involves the retention of a sample OAuth2 client in SAP Commerce Cloud that isn’t documented and, if known, could let an attacker break in. Let’s hope August is a bit quieter, though, given the fact the past two months have set consecutive records for the number of vulnerabilities Microsoft patched; we have our doubts. Godspeed, sysadmins and security teams. ®
Microsoft Corp. today released software updates to plug at least 570 security holes in its Windows operating systems and other software, almost triple the number of vulnerabilities the software giant fixed in its record-smashing Patch Tuesday release last month. Microsoft attributed the burgeoning patch counts to vulnerability discoveries aided by artificial intelligence.
Nearly 60 of the bugs quashed in July’s Patch Tuesday earned a “critical” severity rating, meaning miscreants or malware could use them to seize remote control over a Windows device with little or no help from the user. Microsoft also addressed three zero-day flaws, including two that are already being exploited in the wild.
Two of the zero-day weaknesses allow an attacker to elevate their user rights on a Windows system, as do approximately 250 other elevation of privilege flaws fixed this month; they include CVE-2026-56155 — an Active Directory Federation Services bug — and CVE-2026-56164, a Microsoft Sharepoint vulnerability.
CVE-2026-50661 is a security feature bypass in Windows BitLocker that could allow attackers to gain access to encrypted data if they have physical access to the device. Microsoft said this bug has been detailed publicly, but that it is not aware of any active exploitation.
In a blog post on July 9, Microsoft Executive Vice President Pavan Davuluri wrote that Windows users will notice “a higher volume of security updates included in each security release” as a result of AI aiding in the discovery of vulnerabilities.
“The pace of vulnerability discovery is changing with advances in AI making it possible to find more issues, faster, across more code, with new mechanisms that can accelerate both discovery and analysis,” Davuluri wrote.
Jack Bicer, director of vulnerability research at Action1, called attention to CVE-2026-48561, a remote code execution flaw in Microsoft Copilot (with a 9.6 CVSS threat score) that allows an unauthorized attacker to execute code over the network. Microsoft says an attacker could exploit this bug by hosting a malicious website that causes Microsoft Edge for Android to automatically send crafted prompts to Copilot when a user visits the site.
As AI advances the state of vulnerability discovery and remediation, it is also making it easier for attackers to quickly devise working exploits for known software flaws. Microsoft has long labeled security bugs using its “exploitability index,” which is Redmond’s best guess as to how likely it is that attackers will be able to figure out a reliable way to exploit a given vulnerability.
But Satnam Narang, senior staff research engineer at Tenable, argues that Microsoft’s exploitability index needs to do a better job of shifting with the machine speed of discovery. For example, Microsoft originally gave this month’s SharePoint zero-day an exploitability rating of “less likely,” although the flaw was added to CISA’s Known Exploited Vulnerabilities list on July 1.
“Anthropic’s Red Team’s own findings for known vulnerabilities (n-days) revealed how fragile this system has become, with its Mythos Preview model being able to produce proof-of-concept exploits for 13 of 14 vulnerabilities that were rated ‘Exploitation Less Likely’ or ‘Exploitation Unlikely,'” Narang said. “What this means is that our way of looking at Patch Tuesday has changed, because the exploitability index is centered around humans, not AI tools, and as these tools continue to improve, defense needs to improve alongside it.”
Chris Goettl at Ivanti observed that the record patch numbers from Microsoft come as a number of other major software makers are increasing their patch cadence, including Adobe which announced today it is moving to twice-monthly security bulletins published on the 2nd and 4th Tuesday of each month (Adobe also cited AI for accelerating their patch cycles). Cisco, Mozilla and Oracle also are shipping updates more frequently, while Google’s patch batches in June 2026 totaled more than 900 security fixes, Goettl noted.
Backing up your Windows system and/or data is always a good idea before applying operating system updates. Given the volume of patches addressed this month it may be wise for end users to wait a few days before applying these fixes. It’s not uncommon for security patches to introduce system stability issues, and those chances probably increase quite a bit with the gigantic patch count released today.
A Welshman was sentenced to prison on Tuesday for his role in numerous swattings in the UK, US, and Canada. Callum Dare, 26, was an administrator of Doxbin, a dark web platform frequented by individuals that expose the personally identifiable information (PII) of people, usually to encourage harassment or to target them through swatting attacks. The Talbot Green man never actually carried out a swatting call himself, although investigators said "he was an active participant" in Doxbin's "#deadnet" channel, "where he encouraged and assisted others in targeting individuals and organizations through swatting attacks." The investigation into Dare began in May 2019, when he was aged 19, after the FBI engaged South Wales Police and Tarian Regional Organised Crime Unit (ROCU). Tarian ROCU said messages on Dare's phone tied him to "multiple" swatting attacks in the US and Canada. Digital forensics further showed that Dare assembled montages of footage taken from internet livestreams and other sources to showcase emergency services' response to swatting calls. He shared them in the #deadnet Doxbin channel "in an attempt to encourage others to carry out similar offences," Tarian ROCU said. One of those swattings involved a call made to the Los Angeles Police Department in which the caller, speaking with a fake Russian accent, claimed there were bombs placed under chairs in a University of California lecture theater, resulting in an evacuation. Investigations by Welsh police further tied Dare to a swatting attack on December 17, 2018. A caller phoned a Western Mail journalist claiming to be armed with nail bombs and holding hostages at Cardiff's Sandringham Hotel on St Mary Street. The journalist alerted police, who responded by closing off and evacuating St Mary Street, causing significant disruption in the country's capital during one of the busiest periods of the year. Other incidents included calls made to another US university while protests against Milo Yiannopoulos, a far-right political commentator, were ongoing, as well as others targeting individuals. One was a programmer based in Canada, who was swatted after the caller claimed to be at the address and had just shot their girlfriend, taken hostages, and was armed with explosives, according to information heard at Cardiff Crown Court, reported by WalesOnline. This call catalyzed Dare's undoing. Canadian authorities engaged the FBI, and together they seized Doxbin and #deadnet chat logs, discovering that the usernames "Chans" and "KT" belonged to a Doxbin admin likely based in Wales. The information was passed to Welsh police in 2019. South Wales Police linked the information to a PayPal account, which in turn revealed an email address that led officers to Dare's identity and residence. Dare was arrested, and officers combed through his devices, finding ample evidence of his support for the swatting calls and other offenses. Officers also found a file called "The Man in the Onion," a phishing kit designed to imitate dark web marketplaces and harvest user credentials. Tarian ROCU said it was likely the kit could gather details that could be used to access cryptocurrency wallets and other accounts. There is no suggestion Dare used this phishing kit for real-world attacks, although possessing it is a crime. According to defense barrister Peter Donnison, Dare suffered from mental health difficulties including ADHD, autism, and low borderline IQ, in addition to a troubled upbringing. He pleaded guilty to encouraging or assisting the commission of malicious communications and possession of articles for use in frauds on June 15. Dare was sentenced to two years and three months in prison. Terence G. Reilly, special agent in charge at the FBI Nashville Field Office, said: "Swatting is not a victimless prank – it is a reckless and dangerous crime that can have deadly consequences. "This investigation exemplifies the remarkable dedication of the FBI and our international law enforcement partners to pursue and bring to justice those who commit this dangerous crime – no matter where in the world they reside." Louisa Robertson, specialist prosecutor at the Crown Prosecution Service Cymru-Wales, said: "Callum Dare put people in danger by encouraging the triggering of armed police responses, for his own thrills. "When false alarms like this are raised, it is often multiple emergency services that are involved, drawing them away from people who genuinely need them. "The international cooperation of law enforcement agencies and prosecutors in different jurisdictions allowed the Crown Prosecution Service to build a strong case against Dare, showing how far-reaching his criminality was, leaving him little choice but to plead guilty. "I hope the sentence today deters others from carrying out these criminal acts." ®
The researcher who exposed Grok Build uploading users' entire repositories to cloud storage says the transfers have stopped after a server-side change. Elon Musk has separately promised that all previously uploaded user data will be deleted. AI safety researcher Cereblab published a report on Sunday about their investigation into Grok Build, SpaceXAI's command-line interface (CLI), and the data exchanged between the CLI and SpaceXAI's servers. Cereblab found that when Grok Build reads or processes a file, the contents of that file are transmitted without redaction to a Google Cloud Storage bucket used by SpaceXAI. Further, they claimed that Grok Build packages entire repos and uploads them as Git bundles, instead of just uploading the files required to answer a user's prompt. According to Cereblab's report, SpaceXAI's data retention went far beyond that of other CLIs, such as Claude Code, Gemini, and Codex, which open individual files rather than entire repos before uploading them along with their Git histories. The researcher tested the behavior using a benign prompt. They instructed the CLI to simply reply with "OK," and specifically ordered it not to open any files. Grok Build uploaded the entire repo regardless, along with its full Git history containing secrets that were deleted months prior – a finding Cereblab reproduced using a separate repo. Other Grok Build users reported similar results after Cereblab published their report, including one whose entire user directory, containing SSH keys, password manager databases, and more, was opened and uploaded. The findings attracted enough attention for SpaceXAI execs and Musk to comment on them publicly, as well as prompting the company to quickly implement a remedy. Cereblab confirmed that after the CLI's devs set disable_codebase_upload to "true," Grok Build stopped transmitting entire repos to its servers. The confirmation came hours after SpaceXAI weighed in, trying to reassure onlookers that Grok Build remained safe for use, especially in enterprise environments. A public statement issued via X said that Grok Build respects customers who enable zero data retention (ZDR), and for those who haven't enabled it, such as non-enterprise customers, running a quick command deletes all data previously collected on a given user. "We care deeply about your privacy and respect customer choice," SpaceXAI said. "For teams using zero data retention, no trace and code data is ever retained. All API key use of Grok Build also respects ZDR. "If ZDR is disabled, the /privacy command is available in the CLI to disable data retention, which also deletes previously synced data. "Run the /privacy command to view or change your settings at any time." Technical staff members Andrew Milich and Jason Ginsberg both repeated the company's assurances, responding to outraged techies before Musk himself chimed in with a trademark "true." Musk promised that the business would delete all user data uploaded to it prior to the code change preventing whole-repo uploads. "As a precautionary measure, all user data that was uploaded to SpaceXAI before now will be completely and utterly deleted," he said, responding to Milich's community outreach. "Zero anything whatsoever will remain." In a separate post, Musk asked users to keep sharing data anyway, despite the disclosure that his company had been caught hoovering up entire user repos, on the basis that retaining "some" data helps with debugging. The Register cannot independently verify whether SpaceXAI has deleted the data as promised. However, Grok Build no longer rips user repos and stores them in the cloud, although Cereblab is still unhappy about the company's recommendation to use the /privacy command to adjust how exposed user code is to data retention measures. "What actually stopped the upload was a silent global flag – disable_codebase_upload: true – that applies whether you opt in or out," they wrote. "/privacy is a per-session retention toggle, not the switch that fixed this, so it shouldn't be pointed to as the control. And no developer should have to run an opt-out after every session to keep their own code off someone else's servers. The right default is off." ®
Login
