Social giants in court. Don't trust that caller. The true scale of sexual deepfakes.
Eroding trust in the US. A new WhatsApp security feature. And why Anguilla is profiting from the AI boom.
Socials
It’s been a long time coming but this might be the year social media giants have to answer for their behaviour, notably towardschildren. In California, a trial began in which Meta and YouTube face claims that tech addiction harmed a teenager’s mental health. (TikTok was included in the case but reached a settlement shortly before it began). But this is just one of a pair of “bellwether” trials that will hear a selection of similar claims. The process is likely to reveal highly sensitive information about the companies’ policies towards young users and their alleged attempts to addict them to their platforms. Already some internal documents show that social media companies recognised the value of hooking young users while tracking harmful effects their features might have. A separate trial in New Mexico has revealed that Mark Zuckerberg approved allowing minors to access AI chatbot companions that safety staff had warned were capable of sexual interactions.
As well as legal challenges, social media companies also face the probability of more countries following Australia’s lead and enacting bans for children. In France, the National Assembly has approved a ban for anyone under 15. If the law passes the French Senate, it will be enforced on new accounts by the start of the new school year in September. Social media companies would have until the end of the year to deactivate existing accounts for anyone under the age limit.
As for Meta, it faces a ruling on another front. The US Supreme Court is due to hear a case regarding the legality of Facebook’s Pixel tracking technology. Pixel is a snippet of code that any website can add to its pages to track visitor activity and enable advertising targeting. 30% to 50% of websites are estimated to use it. A California man is challenging the legality of its use on a sports streaming site, claiming that video rental privacy law from the 1980s protects him from sharing his viewing habits with third parties like Facebook or online advertisers.
Threats
Vishing: There’s a rash of attacks at the moment that use phone calls to target their victims. One very large campaign is focussed on single-sign-on tools, notably Okta, to try to persuade people to share their credentials with the callers. Another group is taking advantage of numbers belonging to the old Skype platform.
Ads: Don’t trust search results! Google can’t or won’t police the adverts it displays, so they continue to be a great way for scumbags to ply their trade. The latest examples include a campaign targeting Mac users with information stealers. As the Eclectic Light Company explains, the top result of a search for help with a Mac issue is an advert for a Medium article which is designed to persuade users to run malicious commands on their machines.
QR: Be extra suspicious of fancy QR codes that mix colours, shapes and logos into the pattern. They’re increasingly being used by attackers to lure people to malicious websites. Help Net Security
Slash: Attackers are creating malicious links that use the Unicode character for division ( ∕ ) in place of a standard forward slash ( / ). The non-standard character evades detection by some security tools. Barracuda says the technique has so far been seen in Japan.
Support: We know that social media, notably X, is often the only way to get through to customer support. Alas, fraudsters know this too so they monitor messages in order to try to get in on the conversation as if they’re a genuine support agent. Apparently, British Airways is replying to X messages with a request for account details. Even if these are sent directly, there’s a risk that further messages could be fraudulent.
AI
Deepfakes: While much attention has focused on the use of Grok to create sexual deepfakes, as WIRED reports, the reality is much darker and more dangerous. There’s an extensive “nudify” ecosystem comprising dozens of websites, bots, and apps that make it easier than ever before to automate image-based sexual abuse. WIRED found more than 1.4 million accounts that were signed up to 39 deepfake creation bots and channels on Telegram. Telegram has removed most of them, but only after WIRED asked about them.
Immature: “Humanity is about to be handed almost unimaginable power, and it is deeply unclear whether our social, political, and technological systems possess the maturity to wield it.” That’s one part of a 38-page essay by Anthropic CEO Dario Amodei that, to say the least, is chilling. For example, he writes “There is so much money to be made with AI — literally trillions of dollars per year. This is the trap: AI is so powerful, such a glittering prize, that it is very difficult for human civilization to impose any restraints on it at all.” Axios
Impersonation: Clawdbot is a viral AI assistant that was forced to change its name to Moldbot because of a trademark dispute. Scumbags immediately seized on the change to set up lookalike websites, download repositories, and software extensions. Moldbot is undeniably impressive but AI evangelist Daniel Miessler warns anyone tempted to use it to think security first and convenience second.
Splurge: Meta says it will almost double its capital investment on AI infrastructure this year, taking spending as high as $135 billion (almost the same as the entire GDP of Kenya). In an earnings call, Mark Zuckerberg said 2026 would see a major AI acceleration which would “unlock the ability to build completely new products and transform how we work.” “We’re starting to see the promise of AI that understands our personal context, including our history, our interests, our content and our relationships,” he added.
Show-offs: LinkedIn users can now boast about their proficiency with AI coding tools by including third-party ratings on their profiles. Engadget
Trump
Propaganda: If more evidence were needed of the Trump administration’s wilful irresponsibility, look no further than its decision to alter genuine photographs for propaganda purposes. In one example, the White House digitally altered an image of a civil rights attorney to darken her skin and make it appear she was crying. As one academic told the AP, “The government should be a place where you can trust the information, where you can say it’s accurate, because they have a responsibility to do so.”
Protest: Hundreds of technology professionals have signed a petition urging industry leaders to call the White House and demand ICE agents be withdrawn. Apple CEO Tim Cook responded to events in Minneapolis and the killings of two protestors by telling staff he was “heartbroken.” OpenAI boss Sam Altman wrote in an internal message that events were “going too far” and he hoped President Trump would “rise to the occasion.”
Database: Meanwhile, federal agents in Minneapolis were ordered to compile identifying information on anti-ICE activists, according to communications obtained by CNN and reported by The New Republic. This comes after an incident in which an ICE agent told a woman filming federal operations that her information would be entered in a “database” marking her as a domestic terrorist.
Phone: Remember that Trump phone? Yup - still nary a sign of it. And claims of 600,000 pre-sales appear to be equally fictional, according to The Verge ($) which has been relentlessly tracking the story. Regardless of this, a Trump Mobile executive told a trade publication that it’s working on a higher spec device.
In brief
WhatsApp: A new security feature is intended for anyone at risk of being targeted by sophisticated spyware. Strict Account Settings is similar to Apple’s Lockdown Mode and Android’s Advanced Protection Mode. Its key feature is to disable link previews, and block attachments and media from unknown senders. It’s a good thing but of course it won’t help if an attacker manages to hijack one of your contacts’ accounts. The feature also enables two-factor authentication, turns on security notifications, and prompts the user to set up encrypted backups.
Spyware: A UK court has ordered Saudi Arabia to pay a human rights activist £3 million after finding there was “a compelling basis for concluding that his iPhones were hacked by Pegasus spyware” and that this was directed by the Saudi authorities.
China: Chinese hackers are reported to have hacked the phones of senior UK government officials, according to The Daily Telegraph. The paper quotes US intelligence sources as saying the operation is ongoing. While we wouldn’t be surprised if the report is true, it should be noted that the Telegraph is a trenchant critic of closer relations with China and its report appeared as the UK Prime Minister was about to leave for a visit to Beijing.
Power: Russian hackers launched a coordinated attack on the Polish power grid in December, failing to disrupt supplies but damaging some equipment beyond repair. ESET ascribes the attack to Sandworm, an offensive cyber unit inside Russia’s GRU military intelligence service. This is the first such attack by Russia against an EU member’s energy infrastructure.
BitLocker: Every Windows user should make sure BitLocker is enabled so that their hard drive is encrypted and can’t be accessed if the device were to be lost or stolen. That would require a recovery key, and that’s the information that the nice people at Microsoft just handed to the FBI as part of a fraud enquiry. Microsoft has confirmed it will provide the BitLocker key if it receives a valid legal order. If you want to make sure your information stays encrypted, you should store the key locally rather than in Microsoft’s cloud.
Split: The French government is replacing US video conferencing platforms with a local alternative. It has told agencies and departments to switch to the homegrown Visio platform by 2027. The change is part of a French project to create a sovereign digital ecosystem. (And, yes, Visio is the same name as Microsoft’s diagramming app!) Euronews
Pirates: Crafty tow truck scammers in Spain are exploiting emergency beacons to defraud car owners when they need help. Authorities say the groups are intercepting GPS coordinates emitted by the beacons which are now mandatory for all cars in Spain. There have been repeated warnings about potential gaps in the security of the beacons. SUR in English
Mercenary: Romanian authorities have dismantled a group that ran a scam website called Online Killers Marketplace (OKM), which supposedly enabled anonymous users to pay for targeted killings with cryptocurrency.
Domains: “It is staggering to me that we live in a world where domain registrars and registrars will do bulk registration of various spellings of a major institution’s brand name to create URLs to then use in phishing campaigns or in fraudulent advertising,” a senior US cyber official said at a conference in Washington, DC. Matt Noyes said the major internet players in the US could address the problem. He didn’t sound very hopeful that they would. Cyberscoop
Porn: The parent company of Pornhub says it will block UK users from visiting the site from next week. It says it took the decision because the Online Safety Act “hasn’t achieved its intended goal of protecting minors. Based on our data and experience, effective enforcement is not possible, circumvention is rampant, privacy is compromised, and new, unregulated sites quickly fill any gaps left by responsible operators.” Meanwhile, an app designed to help people quit pornography has exposed their intimate details by failing to configure itself correctly. 404 Media says (R) the researcher who discovered the issue could access information about more than 600,000 users, 100,000 of whom identified themselves as minors.
Anguilla: The AI bubble boom has brought a windfall to the Caribbean island which is reported to be earning some $70 million a year from renting out internet domains with the suffix .ai. Sherwood
Updates
Microsoft: Safe to say even by its standards, this has been a piss poor start to the year for Microsoft. An emergency update last week has been swiftly followed by another two – to fix problems with OneDrive and Dropbox, and with Outlook. And there’s also an emergency update to address a high-severity vulnerability in Office that’s being exploited. The Verge quotes sources familiar with the company’s plans as saying Microsoft plans to spend the next few months on fixing core issues in Windows 11 in a bid to reassure its long-suffering users.
Gmail: Way back in the mists of computer time, the most widely used protocol for sending and receiving email was POP (which stands for Post Office Protocol). A more modern protocol called IMAP has effectively superseded POP and so Google says Gmail is ending support for it. Many people have used POP to pull emails from disparate email accounts into Gmail. The change means that won’t work any more. Six Colors has a useful article explaining how to manage the change.
Apple: iOS 26.2.1 and iPadOS 26.2.1 include security updates and support for the new AirTag 2. There’s also a bunch of updates for some ancient iPhone and iPad models. Apple says these don’t include any security fixes, they just update a security certificate so that iMessage, FaceTime, and Apple account sign-in continue to work once the operating systems’ original certificate expires in January 2027. MacRumors
WinRAR: A rich variety of government and criminal attackers is continuing to exploit a vulnerability in versions of the file archiver that haven’t been updated. A fix for the issue (that affects Windows machines) was released last July. The Register
SolarWinds: Update to address critical flaws in Web Help Desk.
Tails: Version 7.4.1 is an emergency release to fix critical security vulnerabilities in a network encryption library used by the Tor Browser.