|
Another week in the books, and it's time to recap. Buckle up, we got new slop. ICYMI: 404 Media is suing ICE for access to its $2 million spyware contract with Paragon. We filed the lawsuit earlier this week. “404 Media has asked ICE to disclose agency records relating to its contract with a company known for its powerful spyware tool whose potential use in the agency’s ongoing mass-deportation campaign has prompted lawmakers, civil liberties organizations, and immigration groups to express deep concerns over potential civil rights abuses,” the lawsuit says. Supporters helped make this possible, and the best way you can help is by becoming a paid subscriber.
|
|
|
|
|
Lock down your account logins across devices. The YubiKey by Yubico is a physical security key that protects your digital life from phishing attacks. It safeguards against AI-powered cyberattacks, phishing, online identity scams, and account takeovers. Unlike other multi-factor authentication (MFA), it requires a physical tap to verify your login—shutting hackers out, even if they have your password. Millions of users trust YubiKey to protect the personal and business accounts you love and 1,000+ more, including email, social, password managers, productivity tools, developer tools, and more. YubiKeys are securely manufactured in Sweden and programmed in the USA. Still undefeated. The original passkey. For a limited time, get $10 off your order of exactly two keys from the YubiKey 5 Series or Security Key Series using the code “404media10” at checkout. Learn more and get $10 off at yubico.com/404media
|
|
|
|
On the podcast this week: How Paragon sells tech for remotely breaking into phones and reading messages from encrypted chat apps without a target even clicking a link, and a couple of stories about AI “workslop” and the engineers who fix peoples’ vibe coding. In the section for subscribers at the Supporter level, we start with a malicious game on Steam stealing cryptocurrency from a cancer patient, then we talk about Silksong. Listen to the weekly podcast on Apple Podcasts, Spotify, or YouTube.  Image: Secret Service SECRET SERVICE SMS SIM SCAMOn Tuesday, the Secret Service said it “dismantled a network” of “300 co-located SIM servers and 100,000 SIM cards across multiple sites.” The Secret Service suggested that this network posed a threat to the United Nations General Assembly meeting, which was “within 35 miles” of the servers and which it said could have been used to “disable cell phone towers.” The technology used, which can be seen clearly in photos released by the Secret Service, are regularly used by SMS scammers, spammers, and marketers, yes, but the tech is also extremely widely used by ticket scalpers seeking to create lots of Ticketmaster accounts with which to buy tickets. This is off-the-shelf technology that anyone can buy and use.  Image: The White House via Flickr "CLASSIC DOUBLE SPEAK"Multiple Palantir and Flock sources say the companies are spinning a commitment to "democracy" to absolve them of responsibility. "In my eyes, it is the classic double speak," one said. In a blog post published in June, Garrett Langley, the CEO and co-founder of surveillance company Flock, said “We rely on the democratic process, on the individuals that the majority vote for to represent us, to determine what is and is not acceptable in cities and states.” The post explained that the company believes the laws of the country and individual states and municipalities, not the company, should determine the limits of what Flock’s technology can be used for, and came after 404 Media revealed local police were tapping into Flock’s networks of AI-enabled cameras for ICE, and that a sheriff in Texas performed a nationwide search for a woman who self-administered an abortion.  Image: Nick Fewings via Unsplash NEW SLOP JUST DROPPEDA joint study by Stanford University researchers and a workplace performance consulting firm published in the Harvard Business Review details the plight of workers who have to fix their colleagues’ AI-generated “workslop,” which they describe as work content that “masquerades as good work, but lacks the substance to meaningfully advance a given task.” The research, based on a survey of 1,150 workers, is the latest analysis to suggest that the injection of AI tools into the workplace has not resulted in some magic productivity boom and instead has just increased the amount of time that workers say they spend fixing low-quality AI-generated “work.” VERY NORMAL INTERNETYouTube removed a channel that was dedicated to posting AI-generated videos of women being shot in the head following 404 Media’s request for comment. The videos were clearly generated with Google’s new AI video generator tool, Veo, according to a watermark included in the bottom right corner of the videos. The channel, named Woman Shot A.I, started on June 20, 2025. It posted 27 videos, had over 1,000 subscribers, and had more than 175,000 views, according to the channel’s publicly available data.  Screenshot via Pump.fun BRIDGE TOO FARA cancer patient lost $32,000 in crypto after installing a Steam game on his computer containing malware that drained one of his crypto wallets. Raivo Plavnieks is a 26 year old self-described “crypto degen” from Latvia who streams on the site Pump.fun under the name Rastaland. After a seven hour stream on September 20, Plavnieks logged off and cashed out his earnings from the stream. Literally seconds later, someone drained those earnings from his wallet, according to an archive of the livestream and blockchain records reviewed by 404 Media. KEEP READINGResponding to AI ‘Workslop’ Is Killing Productivity and Making Workers Miserable, Chloe writes: “I literally JUST had a 'brainstorming' meeting about this with the execs at one of my workplaces, a design & writing company. They gathered up all the writing and editing staff and asked us to explain why the company should or should not pivot to AI content generation, with all of us recast as rewriters of AI content. Their argument was that fewer and fewer clients want to pay for human writers (probably true) and that, in their experience, AI-written content 'can be good if the prompt is designed properly' (bogus!). It was frustrating trying to explain the problems with AI writing to non-writers. I figured my best argument boiled down to money: it takes me 3-4 times as long to rewrite AI content than to write something original. ChatGPT writes in structurally and grammatically correct sentences, but the logic is almost always bad. And I don't mean just incorrect facts– it will list synonyms within lists, or contradict itself mid-paragraph, or write a sentence that means absolutely nothing. It's a lot more difficult to catch these inhuman, inscrutable errors. The shittiest part is that it's often executives who provide us with the first 'draft' of the AI content we're meant to rewrite, and when we ask to see the data behind some of the draft's claims (so that we can rework it accurately), we find out that the information doesn't exist. So either the exec (not some just lazy worker) knowingly fed false info into the AI, or they just asked it to 'make me a slide deck about an impressive investment firm' and didn't care that none of it was true.”
And Alex Miles also replied, “The refrain I've adopted is, ‘I asked you because I wanted to know what you think. If I wanted to know what ChatGPT would say, I would've asked it myself.’ The goal isn't necessarily to shame someone who brings me LLM output, rather remind them that their thoughts and opinions have value that an LLM can't replace. Also yeah, hopefully prevent them from doing it again since the entire interaction makes more work for everyone involved.”
BEHIND THE BLOGThis is Behind the Blog, where we share our behind-the-scenes thoughts about how a few of our top stories of the week came together. This week, we discuss San Diego and Costa Rica reporting trips. JASON: I’m writing this from sunny Athens, Greece, where I’ve been invited to talk about 404 Media at the IMEDD International Journalism Forum, an annual conference. Over the years, I haven’t been to too many conferences, because honestly it was always too disruptive to the day-to-day journalism and work of managing a team to be able to get away. We’re more than two years into this, but one of the nice things about having this company is that I can mostly get my work done whenever makes sense for me, whether that’s late at night in Los Angeles or early in the morning in Greece. Read the rest of Jason's Behind the Blog, as well as Sam, Emanuel, and Joseph's, by becoming a paid subscriber.
|
