Vercel April 2026 security incident

(vercel.com)

183 points | by colesantiago 2 hours ago

18 comments

  • toddmorey 4 minutes ago
    I've been part of a response team on a security incident and I really feel for them. However, this initial communication is terrible.

    Something happened, we won't say what, but it was severe enough to notify law enforcement. What floors me is the only actionable advice is to "review environment variables". What should a customer even do with that advice? Make sure the variable are still there? How would you know if any of them were exposed or leaked?

    The advice should be to IMMEDIATELY rotate all passwords, access tokens, and any sensitive information shared with Vercel. And then begin to audit access logs, customer data, etc, for unusual activity.

    The only reason to dramatically overpay for the hosting resources they provide is because you expect them to expertly manage security and stability.

    I know there is a huge fog of uncertainly in the early stages of an incident, but it spooks me how intentionally vague they seem to be here about what happened and who has been impacted.

  • nike-17 2 minutes ago
    Incidents like this are a good reminder of how concentrated our single points of failure have become in the modern web ecosystem. I appreciate the transparency in their disclosure so far, but it definitely makes you re-evaluate the risk profile of leaning entirely on fully managed PaaS solutions.
  • swingboy 7 minutes ago
    Is this one of those situations where _a lot_ of customers are affected and the “subset” are just the bigger ones they can’t afford to lose?
  • jtreminio 1 hour ago
    I'm on a macbook pro, Google Chrome 147.0.7727.56.

    Clicking the Vercel logo at the top left of the page hard crashes my Chrome app. Like, immediate crash.

    What an interesting bug.

    • embedding-shape 39 minutes ago
      Huh, curiously; I'm on Arch Linux, crash happens in Google Chrome (147.0.7727.101) for me too, but not in Firefox (149.0.2) nor even in Chromium (147.0.7727.101).

      I find it fun we're all reading a story how Vercel likely is compromised somehow, and managed to reproduce a crash on their webpage, so now we all give it a try. Surely could never backfire :)

      • nozzlegear 18 minutes ago
        Works in Safari too. Sounds like a Google Chrome thing.
    • argee 13 minutes ago
      Happens for me on an M4 Macbook Pro as well, and it happens when you mousedown, so doesn't even require a full click or execution of click listeners/navigation.

      If you remove the href attribute "/home", it stops happening. If you add that attribute to a different link, that link now causes the crash.

    • plexicle 2 minutes ago
      MBP - M4 Max - Chrome 146.0.7680.178.

      No crash.

      Now I don't want to click that "Finish update" button.

    • Malipeddi 25 minutes ago
      Same with Chrome on Windows 11. I opened the vercel home page using the url once after which it stopped crashing when clicking on the logo.
    • burnte 59 minutes ago
      I'm running 147.0.7727.57 and this doesn't happen. Macbook Air M5. VERY interesting.
    • farnulfo 1 hour ago
      Same hard crash on Chrome Windows 11
    • itaintmagic 1 hour ago
      Do you have a chrome://crashes/ entry ?
      • rapfaria 50 minutes ago
        it did add an entry - windows 11, chrome
  • MattIPv4 2 hours ago
    Related: https://news.ycombinator.com/item?id=47824426

    https://x.com/theo/status/2045862972342313374

    > I have reason to believe this is credible.

    https://x.com/theo/status/2045870216555499636

    > Env vars marked as sensitive are safe. Ones NOT marked as sensitive should be rolled out of precaution

    https://x.com/theo/status/2045871215705747965

    > Everything I know about this hack suggests it could happen to any host

    https://x.com/DiffeKey/status/2045813085408051670

    > Vercel has reportedly been breached by ShinyHunters.

    • otterley 1 hour ago
      Who is this “theo” person and why are multiple people quoting him? He seems to have little to say that’s substantive at this point.
      • gordonhart 52 minutes ago
        He’s a tech influencer, probably getting quoted here because he has the biggest reach of people covering this so far.
      • MikeNotThePope 53 minutes ago
        Theo Browne is a reasonably well known YouTuber & YC founder.

        https://t3.gg/

      • reactordev 13 minutes ago
        YT tech vlogger
  • joshmn 7 minutes ago
    Twice in a week for a ShinyHunters story! I'll keep saying this because it's interesting:

    I was in federal prison with Sebastien Raoult of ShinyHunters. We were in the same unit and talked regularly. My takeaway from our conversations about technology/hacking/programming/whatever was that him and his contemporaries have an uncanny level of persistence to phish.

    What Sebastien often described to me was trivial—scanning GitHub for API keys, using sendmail, and your good 'ol phishing page that did auth checks against the real thing. Nothing revolutionary is happening on the attacker's side here.

    What has shifted is development processes and how we think about and perform development. I think we'll only see more and more of these kinds of incidents (not necessarily Vercel-scale) as more and more people-with-ideas have the means to blindly build.

    https://x.com/theo/status/2045862972342313374?s=46

  • OsrsNeedsf2P 2 hours ago
    The lack of details makes me wonder how large this "subset" of users really is
    • bossyTeacher 21 minutes ago
      The lack of details itself is telling enough. Whatever comes out will be no doubt PR sanitised and some bigger clumps of truth won't make it through the PR process.
  • adithyasrin 1 hour ago
    We run on Vercel and I wonder if / how long before we're alerted about a leak. Quick look online suggests environment variables marked as sensitive are ok, but to which extent I wonder.
  • neom 1 hour ago
    https://x.com/theo/status/2045871215705747965 - "Everything I know about this hack suggests it could happen to any host"

    He also suggests in another post that Linear and GitHub could also be pwned?

    Either way, hugops to all the SRE/DevOps out there, seems like it's going to be a busy Sunday for many.

    • phillipcarter 1 hour ago
      I don't know if I'd trust some random programmer-streamer-influencer on anything other than the topic of streamer-influencing.
      • hvb2 59 minutes ago
        The link at the top of the page it to vercel acknowledging it...
    • embedding-shape 1 hour ago
      Based on what, "feels like it"? Claiming that Cloudflare is affected by the same hack has to come from somewhere, but where is that coming from?
      • gruez 1 hour ago
        from his "sources".

        > Here’s what I’ve managed to get from my sources:

        >3. The method of compromise was likely used to hit multiple companies other than Vercel.

        https://x.com/theo/status/2045870216555499636

        To be fair journalists often do this too, eg. "[company] was breached, people within the company claim"

        • eddythompson80 54 minutes ago
          Isn’t he a Vercel evangelist though?
          • troupo 44 minutes ago
            He is "whatever gives me short-term boost in popularity". Including doing 180 turns on whatever he's evangelizing or bashing.
            • eddythompson80 38 minutes ago
              Fair enough. That’s probably a better description from what I’ve seen from him. I remember that arc browser shelling.
            • Barbing 11 minutes ago
              Good for the content but would sponsors be on board long term?
    • recursivegirth 59 minutes ago
      Ah, Theo with his vast insights and connections into everything. That man gets around, and his content is worth it's cost.

      Theo's content boils down to the same boring formula. 1. Whatever buzzword headline is trending at the time 2. Immediate sponsored ad that is supposed to make you sympathize with Theo cause he "vets" his sponsors. 3. The man makes you listen to a "that totally happened" story that he somehow always involved himself personally. 4. Man serves you up an ad for his t3.chat and how it's the greatest thing in the world and how he should be paid more for his infinite wisdom. 5. A rag on Claude or OpenAI (whichever is leading at the time) 6. 5-10 minutes of paraphrasing an article without critical thought or analysis on the video topic.

      I used to enjoy his content when he was still in his Ping era, but it's clear hes drunken the YT marketer kool-aid. I've moved on, his content gets recommend now and again, but I can't entertain his non-sense anymore.

      • rubslopes 11 minutes ago
        I just wanted to chime in and say I think he is knowledgeable; he's not a con. I know you didn't say that, but people might have the impression he doesn't know what he's talking about. He does know, and I've learned quite a lot from him in the past.

        However, since the LLM Cambria explosion, he has become very clickbaity, and his content has become shallow. I don't watch his videos anymore.

      • neom 53 minutes ago
        I don't watch his content, but I felt comfortable posting his link as I believe he's generally considered a reputable guy? His tweets sometimes come up in my for you tab and he seems reasonable and knowledgable generally? Maybe I'm wrong and shouldn't have linked to him as a source.
        • steve_adams_86 44 minutes ago
          He's kind of like an LLM in that his content has the surface texture of something substantial, and sometimes it's backed by substance, yet it's often half-true or totally off the mark too. You'll notice if you're previously acquainted with what he's talking about, otherwise he seems to be as you described.

          I don't think he's a bad guy or that he's trying to be misleading. I suspect he wants his content to actually carry value, but he produces too much for that to be possible. Primarily he's a performer, not a technologist.

          • arabsson 30 minutes ago
            I agree with this comment. YouTube's summarize this video feature has been a godsend when it comes to Theo's videos.
        • threetonesun 43 minutes ago
          Nothing on x.com is reputable at this point.
    • nozzlegear 29 minutes ago
      > @theo: "I have reason to believe this is credible. If you are using Vercel, it’s a good idea to roll your secrets and env vars."

      > @ErdalToprak: "And use your own vps or k3s cluster there’s no reason in 2026 to delegate your infra to a middle man except if you’re at AWS level needs"

      > @theo: "This is still a stupid take"

      lol, okay. Thanks for the insight, Theo, whoever you are.

    • techpression 1 hour ago
      ”Any host” of what? That’s such a non-descriptive statement and clearly not true at face value.
    • rvz 1 hour ago
      I do remember that OpenAI did use Vercel a year ago. They might have likely moved off of it to something better.
  • sreekanth850 43 minutes ago
    Too much of uncontrolled vibecoding?
    • steve1977 39 minutes ago
      While I would agree, unfortunately with JavaScript vibecoding is not even necessary to run into issues.
      • LunaSea 26 minutes ago
        Because Flash apps were so safe.
        • scrollaway 5 minutes ago
          Windows 95 was peak security. (/s)
  • gneray 1 hour ago
    • rubiquity 1 hour ago
      He doesn't work at Vercel but he is the type to never pass up any opportunity to chase clout.
      • threecheese 52 minutes ago
        Almost like that’s his job.

        Hey, I’m with you - I think social media needs to die specifically for this reason. I’m reminded of the term “snake oil” - it’s like the dawn of newspapers again.

  • ofabioroma 1 hour ago
    Time to ipo
  • 0xy 1 hour ago
    This is why you pay a real provider for serious business needs, not an AWS reseller. Next.js is a fundamentally insecure framework, as server components are an anti-pattern full of magic leading to stuff like the below. Given their standards for framework security, it's not hard to believe their business' control plane is just as insecure (and probably built using the same insecure framework).

    Next.js is the new PHP, but worse, since unlike PHP you don't really know what's server side and what's client side anymore. It's all just commingled and handled magically.

    https://aws.amazon.com/security/security-bulletins/rss/aws-2...

    • embedding-shape 1 hour ago
      > Next.js is the new PHP, but worse, since unlike PHP you don't really know what's server side and what's client side anymore. It's all just commingled and handled magically.

      Wasn't unheard of back in the day, that you leaked things via PHP templates, like serializing and adding the whole user object including private details in a Twig template or whatever, it just happened the other way around kind of. This was before a fat frontend and thin backend was the prevalent architecture, many built their "frontends" from templates with just sprinkles of JavaScript back then.

    • sbarre 1 hour ago
      People say "Next.js is the new PHP" because it's the most popular and prominent tooling out there, and so by sheer number of available targets it's the one that comes up the most when things go wrong like this.

      But there are more people trying to secure this framework and the underlying tools than there would be on some obscure framework or something the average company built themselves.

      Also "pay a real provider", what does that mean? Are you again implying that the average company should be responsible for _more_ of their own security in their hosting stack, not less?

      Most companies have _zero_ security engineers.. Using a vertically-integrated hosting company like Vercel (or other similar companies, perhaps with different tech stacks - this opinion has nothing to do with Next or Node) is very likely their best and most secure option based on what they are able to invest in that area.

    • bakugo 23 minutes ago
      Next.js is the polar opposite of PHP, in a way.

      PHP was so simple and easy to understand that anyone with a text editor and some cheap shared hosting could pick it up, but also low level enough that almost nothing was magically done for you. The result was many inexperienced developers making really basic mistakes while implementing essential features that we now take for granted.

      Frameworks like Next.js take the complete opposite approach, they are insanely complex but hide that complexity behind layers and layers of magic, actively discouraging developers from looking behind the curtain, and the result is that even experienced developers end up shooting themselves in the foot by using the magical incantations wrong.

  • nryoo 7 minutes ago
    [dead]
  • jccx70 1 hour ago
    [dead]
  • rvz 1 hour ago
    There is no serious reason to use Vercel, other than for those being locked into the NextJs ecosystem and demo projects.
    • allthetime 48 minutes ago
      I recently got hit by a car on my bike. While I was starting the claim filing process the web portal for ICBC (British Columbia insurance) was acting a little funky / stalling / and then gave me a weird access error. Down at the bottom of the error page was a little grey underlined link that said “vercel”.

      I’m not exactly surprised, but it seems like the unserious, ill-informed and lazy are taking over. There is absolutely zero reason why a large, essential public service should be overspending and running on an unnecessary managed service like vercel… yet, here we are.

  • mikert89 1 hour ago
    Much as I want to rip on vercel, its clear that ai is going to lead to mass security breaches. The attack surface is so large, and ai agents are working around the clock. This is a new normal. Open source software is going to change, companies wont be running random repos off github anymore
    • sph 56 minutes ago
      Your entire recent posting history is "software engineering is over, AI has won."

      What's your agenda here?

      • mikert89 48 minutes ago
        how many recent security breaches have we seen?
        • nozzlegear 32 minutes ago
          How many can unequivocally be attributed to malicious AI?
      • bossyTeacher 51 minutes ago
        Paid by a Sama minion, I bet.
    • goalieca 51 minutes ago
      Slop coding and makeshift sites being thrown up with abandon at breakneck speeds is going to buy me a lot of minivans.
    • tcp_handshaker 1 hour ago
      >> ai is going to lead to mass security breaches.

      Let that be the end of Microsoft. Was forced to use their shitty products for years, by corporate inertia and their free Teams and Azure licenses, first-dose-is-free, curse.

    • lijok 1 hour ago
      ShinyHunters are a phishing group. What does this have to do with AI agents?
      • mikert89 1 hour ago
        Run ai agents around the clock to do hyper targeted fishing
        • cj 1 hour ago
          I feel like humans would be better at hyper targeting.

          AI agents have the benefit of working at scale, probably "better" used for mass targeting.

          • mikert89 57 minutes ago
            this like is saying email marketing is done better if you hand write every email. Thats true, but the hit rate is so low, that you are better off generating 1 million hyper personalized emails and firing them off into the ether
            • mcmcmc 24 minutes ago
              As someone who did the former for a couple years, “better off” is subjective and dependent on your business model, particularly for B2B. It’s a trade off like anything else. You may get more leads, but they may convert at a lower rate. Sending at that scale also increases your risk of email deliverability problems. Trashing your domain has more impacts than you’d think. In smaller, targeted markets it even can damage your business reputation and hurt future sales if done poorly; word gets around.
            • cj 20 minutes ago
              If you’re targeting a million people, I wouldn’t consider that a hyper targeted attack.

              But I get your point.

          • freedomben 1 hour ago
            I disagree. Many humans are phishing in a different language than their native tongue, and LLMs are way better at sounding legit/professional than many of them. The best spear-phishing will still be humans, but AI definitely raises the bar.