[HN Gopher] Human study on AI spear phishing campaigns
       ___________________________________________________________________
        
       Human study on AI spear phishing campaigns
        
       Author : DalasNoin
       Score  : 117 points
       Date   : 2025-01-05 13:40 UTC (9 hours ago)
        
 (HTM) web link (www.lesswrong.com)
 (TXT) w3m dump (www.lesswrong.com)
        
       | TechDebtDevin wrote:
       | Grandma is fkd
        
         | noman-land wrote:
         | We are all grandma.
        
           | pavel_lishin wrote:
           | I've had coworkers, and so has my spouse, who has fallen for
           | the "iTunes gift cards for the CEO" trick. I think _grandma_
           | is no longer an accurate stand-in for a tech-unsavvy person
           | who is vulnerable to spearphishing attempts.
        
             | consumer451 wrote:
             | I would argue that pretty much everyone could be socially
             | engineered into dropping their guard for a moment.
        
             | e40 wrote:
             | I get about 10 emails per week from my "CEO" to pay an
             | invoice. I've even gotten a few text messages. Oddly, the
             | emails never have an attachment. Is this because Google
             | (Workspace account) is removing it?
             | 
             | I've always wondered if it is 10 different orgs doing the
             | campaigns, or the same one. If the same one, why send 10?
        
               | noman-land wrote:
               | This is somehow not considered to be an active warzone
               | when it clearly is. The slightest misstep could ruin your
               | life.
        
               | pavel_lishin wrote:
               | > _I 've always wondered if it is 10 different orgs doing
               | the campaigns, or the same one. If the same one, why send
               | 10?_
               | 
               | My bet is that one criminal group is selling software to
               | enable this, with very similar default settings. Then ten
               | groups by the software, and each one ends up sending you
               | a very similar email.
        
       | terribleperson wrote:
       | This is one of the terrifying, probably already happening threats
       | presented by current LLMs.
       | 
       | Social engineering (and I include spearphishing) has always been
       | powerful and hard to mitigate. Now it can be done automatically
       | at low cost.
        
       | cluckindan wrote:
       | If the study was done with target consent, it might be biased
       | with inflated click-through rates due to the targets expecting
       | benign well-targeted spear-phishing messages.
       | 
       | If it was done without target consent, it would certainly be
       | unethical.
        
         | neom wrote:
         | They got IRB approval. The authors framed the emails as part of
         | a marketing study involving "targeted marketing emails."
        
         | jt2190 wrote:
         | It seems like "the subject clicked a link in an email" is
         | equated to "being phished", but I'm not certain that is a good
         | definition.
        
           | Terr_ wrote:
           | I'm certain that someday I'm going to be dinged on a really
           | shallow kind of work security test because I decided to
           | investigate a link into a sandbox/honeypot environment.
        
             | RaptorJ wrote:
             | These phish testing companies always stick a header
             | (X-PHISH-TEST or some such) on the email so the email
             | server can white-list -- easy to just Outlook blackhole
             | filter anything with that header after you've seen one
             | test.
        
               | Terr_ wrote:
               | What stops an attacker from abusing the same header?
               | 
               | It could be kinda-secure if the header had to have a
               | payload which matched a certain value pre-approved for a
               | time-period. However an insider threat could see the test
               | going on and then launch their own campaign during the
               | validity window.
        
       | joe_the_user wrote:
       | That, after thirty years, email security still depends on the
       | wisdom of individuals not clicking the wrong link, is appalling.
       | 
       | The situation involves institutions happy to opaque links to
       | email as part of their workflow. What could change this? All I
       | can imagine is state regulation but that also is implausible.
        
         | concerndc1tizen wrote:
         | The same is true for operating systems. Why don't they sandbox
         | properly?
         | 
         | We have sandboxing on mobile apps. Why can't we have the same
         | for desktop?
        
           | andersa wrote:
           | After all these years, Microsoft is _finally_ rolling out
           | win32 app isolation, so maybe we are finally on the good
           | path...
        
           | fassssst wrote:
           | Developers initially revolted against Microsoft UWP and Mac
           | App Store.
        
             | wetpaws wrote:
             | Walled garden is not a substitute for security
        
             | Y_Y wrote:
             | Not because they isolated the applications though! Because
             | they were shit, and that's not a requirement.
        
           | joe_the_user wrote:
           | Asking for fully bug free software is nice but unrealistic.
           | Browsers are ostensibly somewhat sandboxed too but there are
           | always new zero-days 'cause browsers are essentially OSes
           | with many moving parts.
           | 
           | However, it reasonable to expect _a single hole_ to be fixed.
           | The  "email hole" has been discussed for decades but here we
           | are.
        
             | makeitdouble wrote:
             | Email is still the running blood of the internet. While we
             | mostly get away with Slack and others for in-group
             | communication, anything going outside, especially to
             | customers, still goes through emails.
             | 
             | At that scale, expecting a core issue to be quickly (or
             | ever) fixed is just unrealistic. I honestly wonder if
             | fundamentally it will ever be fixed, or if instead we get a
             | different communication path to cover the specific use
             | cases we do care about security.
             | 
             | PS: the phone is now 2 century olds, and we sure couldn't
             | solve scamming issues...
        
           | throw10920 wrote:
           | > Why can't we have the same for desktop?
           | 
           | Morally? No reason why, and people are working on it
           | (slowly).
           | 
           | Practically? Because sandboxing breaks lots of things that
           | users and developers like, such as file picking (I hate
           | snaps), and it takes time to reimplement them in a sandbox in
           | the way that people expect them to work. If it requires the
           | developers' cooperation, then it's even slower, because
           | developers have enough APIs to learn as it is.
        
             | bandrami wrote:
             | And to the extent you mitigate some of those user
             | complaints (as flatpak etc. are doing) you are basically
             | re-opening the exact same holes that you developed the
             | sandbox to get away from
        
         | rawgabbit wrote:
         | I blame Microsoft. As a consumer OS its default stance should
         | be no, the user does not intend to grant god permissions to
         | this embedded or external script when they clicked it. Instead
         | the user should have been challenged with a dialog, do you want
         | to install this App and then execute?
        
           | bennythomsson wrote:
           | To which everybody will click yes. They have been conditioned
           | by too much half-baked crap out there that requires it and
           | the need to go on with their lives instead of having tp start
           | investigating things they anyway don't have a clue about (and
           | don't want to, not being IT folks).
        
           | bongodongobob wrote:
           | This is why you don't daily drive a local admin and leave UAC
           | enabled. If you were using an unprivileged account, you'd be
           | getting UAC prompts.
        
         | richardw wrote:
         | Email and web browsing relies on "deny lists" rather than
         | "allow lists". So anything goes but you block bad addresses,
         | rather than nothing until you get
         | permissions/trust/credibility. This helped growth of all the
         | networks but means indefinite whack a mole.
         | 
         | I think (but am not sure) that something using trust networks
         | from the ground up would be better in the long term. Consider
         | anything dodgy until it has built trust relationships.
         | 
         | Eg email servers can't just go for it. You need time to warm up
         | your IP address, use DKIM etc. People can't just friend you on
         | FB without your acceptance so it's a lot safer than email, if
         | still not perfect. A few layers of trust would slow bad actors
         | down significantly.
         | 
         | A trust network wouldn't be binary. Having eg a bunch of spam
         | accounts all trust each other wouldn't help getting into your
         | social or business network.
         | 
         | Thoughts from experts?
        
           | MetaWhirledPeas wrote:
           | > Email and web browsing relies on "deny lists" rather than
           | "allow lists". So anything goes but you block bad addresses,
           | rather than nothing until you get
           | permissions/trust/credibility.
           | 
           | But this is fundamental to an open Internet. Yes going
           | whitelist-only would stop bad actors but it would also hand
           | over the entire internet to the megacorps with no avenue for
           | individual success.
        
             | joe_the_user wrote:
             | Email and browsers shouldn't be glibly equated.
             | 
             | Email as it is presently is a constant opening to phishing
             | and spear fishing. Browser exploits are common too but it's
             | harder (not impossible) to make them personal. And phishing
             | doesn't have to rely on a browser exploit - a fake login
             | page is enough.
             | 
             | It's logical to have a whitelist (or disallow) email links
             | but still allow browsers to follow links.
        
         | ano-ther wrote:
         | I just received a corporate IT security training link. From an
         | external address and with a cryptic link. After a previous
         | training which asked us not to trust external emails
         | (spoofable) especially not with unknown links.
         | 
         | IT wasn't amused when I reported it as phishing attempt.
        
           | nickpinkston wrote:
           | Haha - amazing. I've had the same thought, and I'm sure the
           | scammers have too.
        
         | perching_aix wrote:
         | Technologically, email aliases have been working wonders for me
         | in personal use. No idea if it could be rolled out effectively
         | for nontechnical users at an organizational scale though, even
         | with automation.
         | 
         | It also does little against compromised mailboxes - heck, a
         | sufficiently advanced spear fish might even have better chances
         | if the user misunderstands the security improvements this would
         | provide.
         | 
         | But I think other than this, there's not much else to fix. Some
         | people are malicious, others get compromised. No fixing that.
        
         | rapind wrote:
         | A good start would be ditching HTML in email. Plain text is
         | perfectly suitable for non-marketing emails (and marketing
         | emails are just chaff at this point anyways).
         | 
         | I'll die on this hill.
        
           | MetaWhirledPeas wrote:
           | One word deserves so much blame for the current state of the
           | internet: marketing
        
       | LeftHandPath wrote:
       | They built their phishing emails using data scraped from public
       | profiles. Fascinating.
       | 
       | I have to wonder if, in the near future, we're going to have a
       | much higher perceived cost for online social media usage.
       | Problems we're already seeing:
       | 
       | - AI turning clothed photos into the opposite [0]
       | 
       | - AI mimicking a person's voice, given enough reference material
       | [1]
       | 
       | - Scammers impersonating software engineers in job interviews,
       | after viewing their LinkedIn or GitHub profiles [2]
       | 
       | - Fraudsters using hacked GitHub accounts to trick other
       | developers into downloading/cloning malicious arbitrary code [3]
       | 
       | - AI training on publicly-available text, photo, and video, to
       | the surprise of content creators (but arguably fair use) [4]
       | 
       | - AI spamming github issues to try to claim bug bounties [5]
       | 
       | All of this probably sounds like a "well, duh" to some of the
       | more privacy and security savvy here, but I still think it has
       | created a notable shift from the tech-optimism that ran from
       | 2012-2018 or so. These problems all existed then, too, but with
       | less frequency. Now, it's a full-pressure firehose.
       | 
       | [0]: https://www.wsj.com/politics/policy/teen-deepfake-ai-
       | nudes-b...
       | 
       | [1]: https://www.fcc.gov/consumers/guides/deep-fake-audio-and-
       | vid...
       | 
       | [2]: https://connortumbleson.com/2022/09/19/someone-is-
       | pretending...
       | 
       | [3]: https://it.ucsf.edu/aug-2023-impersonation-attacks-target-
       | gi...
       | 
       | [4]: https://creativecommons.org/2023/02/17/fair-use-training-
       | gen...
       | 
       | [5]: https://daniel.haxx.se/blog/2024/01/02/the-i-in-llm-
       | stands-f...
        
       | hibikir wrote:
       | This lines up well with the success rates I have seen from expert
       | phishers. When I worked at a certain well known company with
       | strong security, a demon called Karla would succeed at
       | spearphishing a bit over 50% of the security team.
       | 
       | AI now means much less skilled people can be as good as she was.
       | Karla as a Service. We are doomed.
        
         | richdougherty wrote:
         | "The cost-effective nature of AI makes it highly plausible
         | we're moving towards an agent vs agent future."
         | 
         | Sounds right. I assume we will all have AI agents triaging our
         | emails trying to protect us.
         | 
         | Maybe we will need AI to help us discern what is really true
         | when we search for or consume information as well. The amount
         | and quality of plausible but fake information is only going to
         | increase.
         | 
         | "However, the possibilities of jailbreaks and prompt injections
         | pose a significant challenge to using language models to
         | prevent phishing."
         | 
         | Gives a hint at the arms race between attack and defense.
        
         | ninkendo wrote:
         | What defines a successful spear phishing? Is it just clicking a
         | link?
         | 
         | My process when I see a sketchy email is to hover over the
         | links to see the domain. Phishing links are obvious to anyone
         | who understands how URLs and DNS works.
         | 
         | But working for a typical enterprise, all links are "helpfully"
         | rewritten to some dumbass phishing detection service, so I can
         | no longer do this.
         | 
         | At my current company I got what I assumed was a phishing
         | email, I hovered over the links, saw they were pointing to some
         | dipshit outlook phishing detection domain, and decided "what
         | the hell, may as well click... may as well see if this phishing
         | detection flags it" [0]...
         | 
         | ... and it turns out it was not only not legit, but it was an
         | internal phishing test email to see whether I'd "fall for" a
         | phishing link.
         | 
         | Note that the test didn't check if I'd, say, enter my
         | credentials into a fraudulent website. It considered me to have
         | failed if I merely clicked a link. A link _to our internal
         | phishing detection service_ because of course I'm not trusted
         | to see the actual link itself (because I'd use that to check
         | the DNS name.)
         | 
         | I guess the threat model is that these phishers have a zero-day
         | browser vulnerability (worth millions on auction sites) and
         | that I'd be instantly owned the moment I clicked an outlook
         | phishing service link, so I failed that.
         | 
         | Also note that this was a "spear phishing" email, so it looked
         | like any normal internal company email (in this case to a
         | confluence page) and had my name on it. So given that it looks
         | nearly identical to other corporate emails, and that you can't
         | actually see the links (they're all rewritten), the takeaway is
         | that you simply cannot use email to click links, ever, in a
         | modern company with typical infosec standards. Ever ever. Zero
         | exceptions.
         | 
         | - [0] My threat model doesn't include "malware installed the
         | moment I click a link, on an up to date browser", because I
         | don't believe spear phishers have those sort of vulnerabilities
         | available to burn, given the millions of dollars that costs.
        
           | bongodongobob wrote:
           | Problem is Outlook now obfuscates the shit out of links,
           | something something safesearch or along those lines. When I
           | hover over a link, I now have no idea where it wants to take
           | me unless I copy and paste it and look through the 500
           | character link to find where it actually wants to take me.
        
         | DoctorOetker wrote:
         | Is this the same Karla as in Fight Club?
        
           | pavel_lishin wrote:
           | I believe that was Marla.
        
       | justinl33 wrote:
       | this research actually demonstrates that AI will reduce the
       | phishing threat long-term, not increase it. Yes, the 50x cost
       | reduction is scary, but it also completely commoditizes the
       | attack vector.
        
         | perching_aix wrote:
         | I'm sorry but I'm not sure I follow. How do you mean that the
         | commoditization of spear fishing will reduce phishing threats
         | long term? To me that implies the exact opposite would happen?
        
       | 101008 wrote:
       | I made a purchase yesterday from Meta (Oculus). A few minutes
       | after payment, I received an email asking to click to confirm it
       | was me.
       | 
       | It came from verify@verification.metamail.com, with
       | alert@nofraud.com cc. All red flags for phishing.
       | 
       | I googled it because it had all the purchase information, so
       | unless a malicious actor infiltrated Meta servers, it has to be
       | right. And it was, after googling a bit. But why do they do such
       | things?i would expect better from Meta.
        
         | tomashubelbauer wrote:
         | I experienced the exact same thing when I bought the Flipper
         | Zero. A "hacker device" and the email communication following
         | the sale being made was straight out of a phishing email
         | campaign book. I don't remember the details, it has been a
         | while, but it was wild how sketchy the emails looked. I hope
         | they have improved the email templates since.
        
           | frizlab wrote:
           | I got way worse. I was fined for leaving an unattended
           | baggage at the train station for a bit. The fine came through
           | an SMS message redirecting to a domain which I had to whois
           | to verify was owned by the train company...
        
         | throw10920 wrote:
         | It's always infuriating getting email from Amazon or my bank
         | "here's signs of potential phishing emails/texts" that doesn't
         | include an exhaustive list of every email address and phone
         | number that that organization will try to contact me from. That
         | should be _table stakes_ when it comes to phishing avoidance,
         | and it 's something that can _only_ be done by the business,
         | not the customer.
         | 
         | Yes, like you say, there's always the chance that someone
         | hijacked an official domain - that's where other things like a
         | formal communication protocol ("we will never ask for your
         | password", "never share 2FA codes", "2FA codes are separate
         | from challenge-response codes used for tech support") and rules
         | of thumb like "don't click on shortened links" come in. Defense
         | in depth is a must, but the list of official addresses should
         | be the _starting point_ and it isn 't.
        
         | ben_w wrote:
         | > i would expect better from Meta
         | 
         | I'm surprised you would expect better.
         | 
         | Everything I hear about their processes, everything I
         | experience as a user, says their software development is all
         | over the place.
         | 
         | Uploading a video on mobile web? I get the "please wait on this
         | site" banner and no sign of progress, never completes. An
         | image? Sometimes it's fine, sometimes it forgets rotation
         | metadata. Default feed? Recommendations for sports teams I
         | don't follow in countries I don't live in. Adverts? So badly
         | targeted that I end up reporting some of them (horror films)
         | for violent content, while even the normal ones are often for
         | things I couldn't get if I wanted to such as a lawyer
         | specialising in giving up a citizenship I never had. Write a
         | comment? Sometimes the whole message is deleted *while I'm
         | typing* for no apparent reason.
         | 
         | Only reason I've even got an account is the network effect. If
         | the company is forced to make the feed available to others, I
         | won't even need this much.
         | 
         | If they stopped caring about quality of their core product,
         | what hope a billing system's verification emails?
        
           | onemoresoop wrote:
           | Yes, but to receive a message that is not from them after a
           | transaction you just did with them is quite bad.
        
         | makeitdouble wrote:
         | Looking at what No Fraud does [0], it sounds like Meta has
         | either spun off the first party hardware store from their usual
         | infra, or straight asked a third party to deal with it, and to
         | insulate their main business they split the email domains.
         | 
         | Most companies are already splitting domains for customer and
         | corporate communication, that's a step in the same direction.
         | 
         | While you're right it sounds fishy as hell, it's also mildly
         | common IMO and understadable, especially when e-commerce is not
         | the main business, and could be a reflection of how anti-
         | phishing provisions are pushing companies to be a lot more
         | protective of the email that comes from their main domain.
         | 
         | [0] https://www.nofraud.com/faq/
        
         | tucnak wrote:
         | You should check your browser extensions!
        
       | 015a wrote:
       | "Look, humans will adapt to the ever-increasing and accelerating
       | nightmares we invent. They always have before. Technology isn't
       | inherently evil, its how it is used that can be evil, its not our
       | fault that we make it so accessible and cheap for evil people to
       | use. No, we can't build safeguards, the efficient market
       | hypothesis leaves no room for that."
        
         | imiric wrote:
         | Mostly accurate, except I would change the last sentence to:
         | 
         | "We take safety very seriously. Look how much safer our SOTA
         | model is based on our completely made up metrics. We will also
         | delay releasing these models to the public until we ensure
         | they're safe for everyone, or just until we need to bump up our
         | valuation, whichever comes first."
        
       | bennythomsson wrote:
       | How did they generate these? If I try with ChatGPT then it
       | refuses, citing a possible violation of their content policy.
       | Even when I tell it that this is for me personally, it knows who
       | I am, and that it's just for a test -- which obviously I could be
       | just pretending, but again, it knows who I am but still refuses.
        
         | qwerty2343242 wrote:
         | You can host open source llm offline.
        
           | ben_w wrote:
           | They team specifically "use AI agents built from GPT-4o and
           | Claude 3.5 Sonnet". The question here is "how did they manage
           | to do so" not "what else can do it with less effort".
           | 
           | As those two are run by companies actively trying to prevent
           | their tools being used nefariously, this is also what it
           | looks like to announce they found an unpatched bug in an
           | LLM's alignment. (Something LessWrong, where this was
           | published, would care about much more than Hacker News).
        
         | a1j9o94 wrote:
         | If you're using ChatGPT directly as opposed to the API, the
         | system prompts could be driving it.
         | 
         | Also, in section 3.6 of the paper, they talk about just
         | switching fishing email, to email helps.
         | 
         | Or said differently, tell it that it's for a marketing email,
         | and it will gladly write personalized outreach
        
       | serviceberry wrote:
       | While I broadly agree with the concerns about using LLMs for
       | "commoditized", large-scale phishing, isn't the study a bit
       | lacking? Specifically, "click through" is a pretty poor metric
       | for success.
       | 
       | If I receive a unique / targeted phishing email, I sure will
       | check it out to understand what's going on and what they're
       | after. That doesn't necessarily mean I'm falling for the actual
       | scam.
        
         | dwood_dev wrote:
         | I hate the InfoSec generated phishing tests.
         | 
         | They all pass DKIM, SPF, etc. Some of them are very convincing.
         | I got dinged for clicking on a convincing one that I was
         | curious about and was 50/50 on it being legit (login from a
         | different IP).
         | 
         | After that, I added an auto delete rule for all the emails that
         | have headers for our phish testing as a service provider.
        
       | Retr0id wrote:
       | It's worth noting that "success" here is getting the target to
       | click a link, and not (for example) handing over personal
       | information or credentials.
        
       | webdevladder wrote:
       | I believe I was the target of employment-flavored spear phishing
       | a few months ago. Could have been a researcher like the OP.
       | 
       | - 3 new email chains from different sources in a couple weeks,
       | all similar inquiries to see if I was interested in work (I
       | wasn't at the time, and I receive these very rarely)
       | 
       | - escalating specificity, all referencing my online presence, the
       | third of which I was thinking about a month later because it hit
       | my interests squarely
       | 
       | - only the third acknowledged my polite declining
       | 
       | - for the third, a month after, the email and website were
       | offline
       | 
       | - the inquiries were quite restrained, having no links, and only
       | asking if I was interested, and followed up tersely with an open
       | door to my declining
       | 
       | I have no idea what's authentic online anymore, and I think it's
       | dangerous to operate your online life with the belief that you
       | can discern malicious written communications with any certainty,
       | without very strong signals like known domains. Even realtime
       | video content is going to be a problem eventually.
       | 
       | I suppose we'll continue to see VPN sponsorships prop up a
       | disproportionate share of the creator economy.
       | 
       | In other news Google routed my mom to a misleading passport
       | renewal service. She didn't know to look for .gov. Oh well.
        
       ___________________________________________________________________
       (page generated 2025-01-05 23:00 UTC)