Copilot edited an ad into my PR

(notes.zachmanson.com)

304 points | by pavo-etc 2 hours ago

40 comments

  • anton-g 1 hour ago
    • jruohonen 39 minutes ago
      Interesting indeed. I wonder how long GitHub as a platform will be there as a viable option. Anyone who remembers SourceForge?
      • Brosper 3 minutes ago
        It's baked in literally into every coding tutorial and is kind of industry standard, like JIRA. Maybe it's just an experiment at this moment.
      • marcus_holmes 30 minutes ago
        I believe Codeberg is the new hotness
        • steve1977 11 minutes ago
          It is, but Codeberg is only for free and open source projects.
      • raincole 15 minutes ago
        A few decades? Its competitors are not magically immune to this kind of spam.
        • jruohonen 13 minutes ago
          > Its competitors are not magically immune to this kind of spam.

          Sure; a platform is a platform is a platform. As for predictions, it is interesting to see whether self-hosting and smaller self-managed infrastructures will gain more traction again.

    • anton-g 29 minutes ago
      Looks like there's a comment added by Copilot before any of these "tips" as well, so pretty sure this originates from Copilot and not Raycast: https://github.com/search?q=%22START+COPILOT+CODING+AGENT+TI...
  • khvirabyan 1 hour ago
    Just thinking, could it be that your coworker used Raycast to spin up a codex to review and fix the typo on the PR? And that comment was added by Raycast?
  • WD-42 1 hour ago
    Why is copilot doing this? If they wanted to show ads couldn’t they… just show ads? Or is GitHub such a house of cards at this point that editing pr descriptions is the only way without risking another 9 of downtime?
    • politelemon 14 minutes ago
      https://news.ycombinator.com/item?id=47570820

      I think this is a ray cast issue, looking at these links. It appears on gitlab too, which is enough for me.

    • flogy 1 hour ago
      Are we sure this actually is originating from MS Copilot itself? Technically I believe it would be possible to smuggle ads into PRs using prompt injection too.
    • oefrha 1 hour ago
      If they show the ad on github.com, agents accessing the PR using (an outdated, ad-free version of) gh CLI won’t see it. /s

      (That said I’m rather skeptical of this and would like to see more details of the process that produced this, and proof.)

      Edit: Just noticed this official GitHub blog post from last month advertising Raycast, making this story a lot more believable: https://github.blog/changelog/2026-02-17-assign-issues-to-co...

      • MattGaiser 59 minutes ago
        It could simply be something in the Raycast integraton?
        • oefrha 52 minutes ago
          I said it’s more believable than GitHub randomly advertising a non-GitHub product (my initial read of the situation, which seemed highly unlikely).
  • post_below 1 hour ago
    Assuming this isn't a hoax, this seems like a huge, probably unintentional, mistake by MS.

    If they genuinely implemented something like this, whatever they made from new customers via ads couldn't possibly make up for the loss of good faith with developers and businesses.

    I suppose if it's real we'll see more reports soon, and maybe a mea culpa.

    • altairprime 1 hour ago
      That’s a really tasteful Juno Mail footer implementation for a mistake. If the AI self-invented it on a lark, good job, but it reads very strongly like someone intended it.
    • chrismorgan 1 hour ago
      How could you implement something like this by accident?
      • rhet0rica 36 minutes ago
        That's a good question! I'm sure we'll find out eventually.

        z Quickly spin up Hacker News comments from anywhere on your macOS or Windows machine with a lobotomy.

      • sheept 1 hour ago
        One feasible scenario could be that they are working on/experimenting with ads, and it was put behind a feature flag, but for whatever reason it was inadvertently ignored
      • bigyabai 1 hour ago
        LLMs aren't known for being super deterministic.
        • mathieudombrock 14 minutes ago
          LLMs are determistic. Just like everything else computers are capable of doing.

          Commercial front-ends just hide the random seed parameters.

    • ccppurcell 43 minutes ago
      Not a hoax, you can search GitHub prs for this string and find many hits.
    • tossandthrow 1 hour ago
      It is likely not a hoax and likely very intentional.

      If you look at the positioning, someone has definitely justified that this is benign and a reasonable place to have an ad added in.

    • goodusername 1 hour ago
      Yeah, would be good to have confirmation that this happened to others as well.

      But it really seems like an own goal if true.

  • nialse 2 hours ago
    Microsoft injecting permanent ads in PRs? Has this been independently confirmed?

    Brought to you by Carl’s Jr.

    • longislandguido 1 hour ago
      > Brought to you by Carl’s Jr.

      I'm reminded of Jay Mohr's legendary take some years back on the creepy Carl's Jr. commercials:

      https://www.youtube.com/watch?v=OJlYRS2Vqkw

    • blitzar 30 minutes ago
      Todays independent confirmation is brought to you by Microsoft — Empowering every person and every organization on the planet to achieve more.
  • pinkmuffinere 1 hour ago
    I think they want the free advertisement, like Apple with its “sent from iPhone” addendums. But “sent from iPhone” is sometimes useful, and significantly shorter. If they just left it at “edited with copilot” I think it would be tolerable
    • supernes 4 minutes ago
      "Sent from iPhone" doesn't contain a call to action, and doesn't exalt the features of the product.
    • politelemon 12 minutes ago
      > But “sent from iPhone” is sometimes useful,

      No, it is still an advert, and not useful in the least.

    • silisili 1 hour ago
      That's exactly where my mind went. It's zero percent more insulting to me than 'sent from my iPhone.'

      If you don't want copilot garbage in your PRs, maybe don't use copilot to create or edit them?

    • computomatic 35 minutes ago
      I don't think the issue is the sign-off so much as that an existing PR was edited. Claude Code signs off when creating PRs and nobody seems bothered. But it won't edit an existing PR, and it won't sign off if I simply ask it not too (which I've automated). Editing any PR it touches - including one authored by someone else - is downright rude.
      • marcus_holmes 28 minutes ago
        > Claude Code signs off when creating PRs and nobody seems bothered

        Not only unbothered, but genuinely appreciative of the notification.

    • winrid 1 hour ago
      It already does that, too, with the co-author
      • pavo-etc 15 minutes ago
        I would argue that is a net positive, it is valuable to know if a language model was involved enough to be committing itself.
  • ex-aws-dude 1 hour ago
    How long before the LLM makes sponsored decisions in the actual implementation?

    "It looks like the user wants to add a database, I've gone ahead and implemented the database using today's sponsor: MongoDB"

    • tossandthrow 1 hour ago
      Likely already happening.
      • nubinetwork 23 minutes ago
        To be fair, Gemini did try to get me to buy some nucleo144s recently...

        (sure, I was working on something embedded, and asked for a recommendation, but it seemed quite intent that it wanted me to use that specific board)

  • simonw 1 hour ago
    Which Copilot was this? There are a bunch of different products that share that name now.
    • SchemaLoad 1 hour ago
      Microsoft has had a lot of naming blunders in the past but this has to be their worst. Copilot is currently, a tool to review PRs on github, the new name for windows cortana, the new name for microsoft office, a new version of windows laptop/pc, a plugin for VS code that can use many models, and probably a number of other things. None of these products/features have any relation to each other.

      So if someone says they use Copilot that could mean anything from they use Word, to they use Claude in VS Code.

      • protocolture 1 hour ago
        >Microsoft has had a lot of naming blunders in the past but this has to be their worst.

        Nah I still rate "Windows App" the Windows App that lets you remotely access Windows Apps. I hate it to death, its like a black hole that sucks all meaning from conversations about it.

        • ValentineC 56 minutes ago
          "Microsoft Remote Desktop" was such a good and distinct name. RIP.
      • hsbauauvhabzb 1 hour ago
        It’s probably a useful feature: if it’s named copilot, assume it’s slop and avoid it.
    • pavo-etc 21 minutes ago
      funny enough I have a page just for tracking this also https://notes.zachmanson.com/microsoft-product-names/
  • napo 1 hour ago
    I wonder if 1) the PR was created using Raycast and this is the model signing its PR, or 2) if there was some prompt injection done at some point.

    Either of these options would still be bad, but here the author suggests that it's just copilot that now just injects ads in its output.

    • pavo-etc 19 minutes ago
      I don't know how Raycast could run on the GitHub servers, but a third potion could be dataset poisoning. Hostile raycast advertising campaign
  • bryanhogan 51 minutes ago
    Whatever the reason for the inclusion was here, the general problem is much bigger. People / companies / products can influence the direction of AI answers to put them in a better light and to be recommended more often. This isn't limited to just products even.
    • SV_BubbleTime 48 minutes ago
      If not on the surface, we’re all deep down aware that an initial era of an advertising-free new technology is once again almost over.

      See you on neural links before “sponsored thoughts”.

      • bryanhogan 8 minutes ago
        It's already over, the problem is the missing transparency. With an LLM you have no idea what influenced the answer, and there is no good way to show it to the user.
  • pabrams 1 hour ago
    Why are you "summoning copilot" to correct a typo?
  • gherkinnn 51 minutes ago
    Obnoxious ads in LLM output was my only 2026 prediction. But I expected OpenAI to get there first and wasn't sure whether the AI companies would first add traditional ad boxes or go straight for blighted responses.
  • rmnclmnt 21 minutes ago
    Isn’t this more of a Raycast issue (apparently an agentic ai service) instead of GH Copilot itself?
  • logicallee 4 minutes ago
  • turtleyacht 1 hour ago
    Do you drive by a billboard that reads

      Does advertising work?
      Just did!
    
    Raycast is an application launcher thing:

    https://en.wikipedia.org/wiki/Raycast_(software)

    Ray casting, however, is different:

    https://en.wikipedia.org/wiki/Ray_casting

  • pants2 1 hour ago
    Was Raycast bought by GitHub or something? Why would it be advertising for Raycast?

    Brought to you by Wendy's.

  • isoprophlex 18 minutes ago
    Satya "please don't say slop" Nadella eat your heart out. Magnificent amounts of value are truly being added by this tech.

    I'll add: it doesnt really matter if this was the integration dumbly appending a message or the llm inserting the ad. Judging by the response to this submission, sneaky ad slop is now firmly inside the overton window, so for MS it doesn't make sense NOT to do it.

  • upmostly 10 minutes ago
    Isn't this the same as

    "Sent from my iPhone"?

  • raincole 57 minutes ago
    Is Raycast even a product of Microsoft? If not, are we witnessing the first large scale prompt injection abuse?
  • oakpond 1 hour ago
    I notice this kind of "Sent from iPhone"-type spam with other AI tools too. It's awful.
  • hexasquid 1 hour ago
    I'm so tired of what initially looks like a perfect normal communication between two people, only to find that some third party has inserted itself like a parasite to exploit and extract human attention. That's why I use our sponsor, nord vpn ...
  • Surac 1 hour ago
    as a non native speaker here please explain the meaning of PR to me.
    • hsbauauvhabzb 1 hour ago
      Pull request, which is a request to merge changes in a git repository.

      Or (not in this case) public relations , which is an interface with how the public views your product, service or company. In this case, copilot adding advertising into git pull requests is bad public relations for Microsoft, but the article author is referring to pull request as PR

  • idkwhatimdoing2 1 hour ago
    Its like microsoft wants to be google, except its very intrusive.

    time is money, save both. try ramp.

  • charcircuit 1 hour ago
    This looks like an ad for only Raycast which does not appear to be affiliated with Microsoft or GitHub at all so blaming Copilot or GitHub here is not justified.

    Edit: The link in the promotion goes to https://docs.github.com/en/copilot/how-tos/use-copilot-agent...

    Which does show that this is affiliated with GitHub unlike what I thought. There are no mentions of this string in a code repository on GitHub (including the Raycast copilot extention).

  • hsbauauvhabzb 1 hour ago
    It was only a matter of time.

    Sent by my iPhone using tapatalk

  • martianlantern 1 hour ago
    Why are they doing this?
  • vcryan 57 minutes ago
    I'm not a fan of LLM's injecting themselves into PR/commit content. If you use multiple models, basically whichever one is operating git gets all the credit. But, even if you wrote all the code yourself, and just submitted the PR with Claude Code (or whatever) it would attempt to take credit for the changes.

    I currently have rules in all of my skill files forbidding models from advertising themselves or taking credit.

  • MattGaiser 1 hour ago
    Post the trajectory if this is real.
    • gpvos 24 minutes ago
      What do you mean with trajectory? Also, a simple github search will show you many hits for the Raycast text, proving that this is quite real.
      • MattGaiser 23 minutes ago
        The path of reasoning the agent took that led it to generate the output. The GitHub search bits got posted after my comment, so while it is clearly real, it just seems injected by Raycast.
        • pavo-etc 15 minutes ago
          This is real. I do not have access to the path of reasoning, this ran through the GitHub copilot app which does not grant you access to the chain of thought.
  • iomer 57 minutes ago
    crappy much. wow.
  • anshumankmr 1 hour ago
  • dinakernel 1 hour ago
    Seriously? Dont they want their system to succeed? I cant think of a better way of alienating the target customer than this.
  • with 1 hour ago
    Everyone is doing this now. Granted, on Codex / Claude Code, you can disable it, it’s not the default to have it disabled. For some reason on Cursor, they keep shoving the “Made with Cursor” into my PR description despite me disabling attribution, which looks really stupid on a work PR.

    I’m so tired of all this BS. Why did this become normal? and how do we not read this as cheap advertising?

    • annie511266728 55 minutes ago
      I think people read it as cheap advertising because a PR isn't really the tool's output, it's team communication.

      A little "made with X" in your own draft is one thing. Putting branding into a PR your coworkers have to read is another.

  • daemin 1 hour ago
    Using a LLM to fix a spelling mistake is retardedly lazy.

    Presumably they used a free version of the LLM, therefore it is completely understandable that it inserted a snippet of text advertising its use into the output. I mean using a free email provider also adds a line of text to the end of every email advertising the service by default - "Sent from iPhone" etc.

    • onion2k 1 hour ago
      Using a LLM to fix a spelling mistake is retardedly lazy.

      If you do it manually, sure.

      If you have an agent watching for code changes and automatically opening PRs for small fixes that don't need a human-in-the-loop except for approving the change, it's the opposite of lazy. It eliminately all those tedious 1 point stories and let's the team focus on higher value work that actually needs a person to think about it.

      Given time all small changes will be done this way, and eventually there won't be a person reviewing them.

      • pabrams 1 hour ago
        That scenario doesn't require any explicit "summoning", and if there's a human in the loop approving the change, certainly they can fix the typo themself.
      • ex-aws-dude 1 hour ago
        Sounds like a great use of energy and tokens, not overkill at all

        In fact I don't even use Ctrl + F anymore and instead just use Claude for all my searches

    • LeoPanthera 1 hour ago
      [flagged]
  • GN0515 1 hour ago
    But... why?
  • minsung0830 21 minutes ago
    [dead]
  • claytonia 59 minutes ago
    [dead]
  • treysu 30 minutes ago
    [dead]
  • winna 56 minutes ago
    [dead]
  • 10keane 1 hour ago
    [dead]
  • ookblah 58 minutes ago
    maybe every PR should be run through 2 other llms so they just remove the ads of competitors (or i guess you'll end up with all 3) /s