Sorry we're so late everyone, between being sick, technical issues with the PC, and other commitments, we're finally here. Some of this news still may be of value, and we're glad to still put it out. Also, starting with this podcast, we're using transcripts. They may not be perfect, but hope they help people. Let me know if it is usable and whether they should be left on. Welcome to Sans, episode 38. This will cover the newsletter that was released on January 13, 2026. Sounds like a group we've talked about is back in the news as they're the top story in "top of the news." We have an Instagram story that we saw through Malware Bytes, but we decided not to blog it. We'll talk about it anyway. We've got some Privacy Protection news from California, it looks like we've got good news from Spanish authorities and we've got a whole lot more. If you would like to read the newsletter, please use this link to do so. Here is what is in the top of the news. Salt Typhoon Threat Actors Reportedly Responsible for New Congressional Email Breach Spanish Energy Company and Supplier Disclose Data Breach LLMs and Healthcare: ChatGPT, Claude, and Google Overviews What do you think about the LLM story which is item 3 of the top of the news section? Here's what is in the rest of the news. LLM APIs Targeted by Threat Actors and Gray-Hat Hackers Instagram Password Reset Emails are Unrelated to Alleged Data Breach BreachForums Member Data Leaked California Privacy Protection Agency Fines Texas Firm for Failing to Register as a Data Broker with the State Spanish Authorities Arrest 34 in Connection with Cyber Crime Network Printing Error Prompts Recall of Nearly 13,000 Recent Irish Passports CISA Retires 10 Emergency Directives This can't be good in regards to the Irish Passport story. Comment on the newsletter by leaving me your thoughts. Contact info is within the program. Thanks so much for listening!
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
The opinions expressed are those of the show hosts and may not necessarily be of any company in which the show hosts may represent. Hello, folks! Welcome to SANS, episode number 38. This... will be the podcast for SANS. This is the News
Bites. This is annotated news update from the leader in information, security, training, certification, and research. This program will cover the newsletter from January 13th, 2026. So, how is everybody doing? We had a bit of a mild setback when it came to my health, and I know that I appreciate
each and every one of you who might drop by and listen to these programs along Jared's Technology Podcast Network. is Volume 28, Number 3. So, this is going to be a very interesting newsletter. And as we always do, if we feel it's necessary,
we will take from the editors. But, we know that sometimes we can just do it by ourselves. Beast! off of the content itself. But we also tell you who is writing so that you know the difference between me talking and the newsletter. So I wrote,
we've got interesting news coming out of this episode of SANS. It covers the newsletter that was issued on January 13th, 2026. We've got a group. that seems to be back in the news. And, of course, it can't be good since it's item one of the section of the newsletter that is titled Top of the News. Since I'm giving you that much,
what is that group? You don't have long to guess because it'll be in the table of contents, which is part of this file. We've got LLM news, especially when it comes to healthcare, as well as LLM news when it comes to it being attacked. We saw an
Instagram story we decided. not to blog, that they'll cover, and there's plenty more. This file is transcribed by Jared Reimer in Woodland Hills, California. If you want a copy, email tech, that's T -E -C -H, at M -E -N -V -I dot org, and I'll take care of it. Let me know what format and size of file you want. Per usual,
we break up each section on a new Braille page and number each item both in the table of contents and the items themselves so you can find them easily. So we go back to numbering each item one through whatever. And, of course, after the particulars of top of the news, it will go to
a new Braille page regardless of code and give you the rest of the week's news. We are using the new UEB page numbering for chapter and page as seen in some books, although I never saw that on any book until... the Unified English Braille course book itself, but I'm using this format
for podcast number, and I will use it as a podcast number and a page number. So, there. You go. Now your time is up in regards to the answer
for today's trivia. The table of contents now follows. Top of the news. Item number one. Salt Typhoon threat actors. Reportedly responsible for new congressional email breach. Alright, so number one is Salt Typhoon for the threat
actor's name. Who would have thought that one? I'm curious. You can email or iMessage, text or WhatsApp. That's 804 -442 -6975. Or call and leave a comment on the comment line at 888 -405 -7524 or 818 -527 -4754. If you do get me, please
talk to me. And do leave a voicemail. Leaving a blank voicemail may not yield any response. Because I don't know who you are or what specifically the number calling wants. So please make sure that you do leave a verbal voicemail. Okay? Number two, Spanish energy company and supplier discloses
data breach. And number three, LLMs and healthcare. ChatGPT, Claude, and Google Overviews. And maybe it might be imperative in that file to talk very specifically about my use of ChatGPT because I have been a little bit... Under the weather,
although I'm doing a lot better at the time I'm recording this, and I'm sure that I will be fine. I'm sure it was either a stomach bug or a combination of stress and fatty foods like a burger and fries and other combinations like ChatGPT said it could be. We just don't know. But... I'll talk about that then. And of course, you'll probably hear
me talk about it on various other places, including TSB and on the RCL program if they want, if they happen to bring up AI. But, more importantly, we'll talk about it on the Throwback Saturday Night program. So, do make sure you check that
out. It'll also be on the Saturday show of the Independent Artist Spotlight, which we are broadcasting at the same time I'm taping this, which is Saturday, January 24th. Speaking of LLMs, we're going to move now to the rest of the table of contents with item one, as there's more LLMs. LLM APIs targeted by threat actors and gray hat hackers.
Number two, Instagram password reset emails are unrelated to... alleged data breach. I saw this in their newsletter and I decided not to blog it. There's really nothing there. I know. Number three, breach forum's member data leaked. That's always great news. I love that. Number four, California Privacy Protection Agency finds Texas
firm for failing to register as a data broker with the state. And this goes back to last episode. I forget the exact item number of this one. I think it was item one of that particular podcast where we talked about this idea where Californians can register. with this particular entity, and
that should, in theory, stop our data from being collected. Item 5. Spanish authorities arrest 34 in connection with Cybercrime Network. 6. Printing error. prompts recall of nearly 13 ,000 recent Irish passports, and that's not good. And finally, number seven, CISA retires 10 emergency
directives, and that is a good sign because they're not needed anymore, notating that they were emergency directives So that is the table of contents. That's the answer to your trivia question. It is Salt Typhoon. If you guessed it, great. If you didn't, learn along. We will continue in
just a moment with the newsletter as we start with item number one of Top of the News. This will be program number 38 of SANS, a Jared Reimer Network podcast. Since we played trivia earlier,
let's go ahead and start with our top of the news item. And, uh... This is item number one. It looks like I didn't put item number one here. But that is okay. You can follow along if you have Braille. Assault Typhoon. They have never been good. We've talked about them since the
very beginning on the security box. Well before I thought of this sans news... news thing as a full podcast set could have done this with all of that coverage then because they talked about it then too but they are back assault typhoon threat actors reportedly responsible for new congressional email breach and I think before
I really get into this, I want to ask a question. And I really think that this is a serious question. If this was Congress, what were they using? Or are they using? I don't... exactly understand
why we're still having issues with getting into email. I don't have anybody getting into mine that I am aware of, and I say that loosely because email can be gotten into. So I'm not going to say that it doesn't happen. But if you don't use the right tools, then of course you're going to get hacked into. But I'm wondering what's being used. Because email's gotta be saved for...
What is it? Many, many years? Like... OMG. Not necessary. Once you respond to something, delete it. There's no need to keep it around. Here's what is... and we'll go through this together, ladies and gents. Cyber threat actors with ties
to China's government have once again infiltrated email systems, compromising U .S. congressional staff members' communications. The intrusion was detected in December 2025 and was first reported reported by the Financial Times, which said the perpetrators are Salt Typhoon, a threat actor
group that gained notoriety for breaching telecommunications systems around the world in 2024. As I said, we covered them on TSB. And they were on Sands then, too. Unfortunately, I did not do Sands then, so this can be looked into. I'm sure you
can contact Sands and ask them, you know, hey, what newsletters did you guys cover Salt Typhoon in? I bet they could help you out. Officials are investigating the breach, which reportedly affects email systems used by congressional staff working on House National Security Committees. Oh, great. So, uh, this can't necessarily be
good. Is that right? So, Salt Typhoon is back again. Is anyone who is hearing this surprised by this news? Especially those who participate on TSB? Dukes writes, It's not surprising that the U .S. Congress is and continues to be a target of nation states. What's left to be determined
is whether the Attack used existing Salt Typhoon TTPs or changed them up? And that, of course, we don't know. And I suppose if I do spot something, I can write something up. If they were known TTPs, then that's a black eye for Congress and
their security team. What security team? Hopefully, that information becomes available soon to protect other organizations. Our guy, Neely, writes, These accounts are targeted as they are typically less hardened environments. While truly... email
is not present, there is enough supporting information which rounds out open source investigations nicely. Make sure that you are considering the security of not only your mainstream email systems, but also of staff, contractor, and supporting services.
An incident can quickly offset the cost of not providing someone with a corporate email account. Email security options to ensure protections are in place. Commensurate with the operation processed. As I wrote, tell me if you've seen
this before, and yes, I did number everything else. I guess I forgot to item number one. That's okay. Folks, this is a huge problem. I still want to know what they're using. Outlook? Microsoft 365? I mean, this is absolutely crazy. I don't...
I don't quite get what they're doing here. This can't necessarily be good. And this isn't going to be the first time. So, time for them to go figure that crap out. Because... That is enough
of that. Coming up, ladies and gents, we're going to move on to our next item. Spanish Energy Company and Supplier Disclose Data Breach. Well, you know... They come a dime a dozen these days, data breaches do. Let's see what this one covers
as the newsletter continues. I'm Jared Reimer. This is Jared's Technology Podcast Network. Alright folks, welcome back to the program. I'm Jared Reimer. Well, I know some time has passed and
people are probably wondering, what happened? Short version, I got sick. But I'm feeling a lot better now, and we are going to push on. We're just going to go the best we can, and I can't rush it. As a doctor would say, don't push yourself too hard. Just do what you can, even if it's just one AI tool. I bet my doctor would
say the same thing, too, if I was telling him, yeah, I'm feeling better today, but I want to do all of this stuff. He'd probably say, no, probably not a good idea to do all of that at once, unless you're up to it. So I guess we'll see how this goes. No judgment, right? Item two of the particular SANS News Bites we are working on is this. Spanish energy company and supplier
disclose data breach. Spanish energy company Endesa and its supplier, NRJ XXI, I don't want to say 21 because I don't want to assume that they mean a number, so I'll just read it as letters, are notifying customers that account information
was compromised in a cybersecurity incident. is one of Spain's largest energy utilities, providing gas and electricity services to more than 10 million customers in Spain and Portugal. Well, that's quite a very large amount of people. It,
quote, detected a security incident that has allowed unauthorized and illegitimate access to its commercial platform. That ends the quote.
The compromised information includes, with Endesa Energe and possibly payment details. And that
ends that quote. Endesa has notified Spain's data protection agency and other relevant authorities. Sacks. A once -in -a -while contributor and who was mentioned in several newsletters writes, This looks like a customer data breach, but from a utility engineering perspective, it is also
a reminder that the commercial platform is part of the operational ecosystem. The grid can be stable and well operated, yet attackers who gain access to IT systems can still degrade utility operations without ever touching control systems.
The stolen data has follow -on consequences. and enables convincing impersonation of customers contractors and even internal staff that credibility is exactly what adversaries need to move from fraud into deeper access especially in environments
where remote support and vendor connectivity are normal. This is not evidence that grid operations were directly at risk. It is evidence that an attacker reached a system boundary that should be treated as reliability relevant. Engineers
should view incidents like this as a prompt to strengthen segmentation, identity controls, and vendor access governance. Those are not just IT best practices. They are part of resilience engineering. that I did not deal with the IT
aspects of things, but you'll understand it. As I write in the next paragraph, stop me if Neely's comments may be somewhat familiar. It may be written a bit differently, but he writes, the breach didn't impact service, in parentheses, gas power. delivery to customers, and Endesa
is notifying affected customers directly. While Endesa is claiming there is no attempted use of the purloined data, threat actors appear to have 20 million records, 1 TB, by the way, that's in parentheses, of Endesa customer database data
for sale to a single exclusive buyer. Expect Endesa to implement enhanced security measures after admitting existing security fell short of expectations. And, oh boy, what do I think?
If this is the only thing I record today, which I think it may not be, because I'm actually doing quite well, thank you very much, I'll say this. Endesa, you better... come clean. And you better tell your customers exactly what went wrong when
you know it. Not six months to a year from now, but as soon as it's feasible for you to do it. Since you're already starting the process, if it were me and I find out new information, I would continue to send out notices so that your customers are aware of what is going on before it hits the press. Now that would be what I would
do. But I'm not Endesa. And I'm not any other company who's been targeted in this list. I think if I were ever breached, and it got out, and it hit sans news bites, the way that I would do things is notify people as soon as possible. If I got pushed back by police, I would say, the problem is that if I don't notify my customers,
then it gets out six months to a year later, and that's bad practice. This is what we must do, and we can do it in such a way where it won't hurt the investigation. I .e., your information was taken, we don't know by whom, there's an ongoing investigation, but you need to know that
this is what's happening. When it's safe for us to give you more information about what specifically was taken, then we'll do that. At least start there. Then when you are released from the police and their investigation, you can then come clean to say what was taken and how you are going to prevent it. But at least giving a heads up that
information was taken from you is a first start. And the fact that an ongoing investigation is preventing you from giving details, you can mention that. But be as forthcoming as you can and continue to provide updates. We've got two LLM stories coming up. One before the rest of the news, and
one after we start the rest of the news. And as we do, and we've started to do in these podcasts, we'll play a track between each. But first, we will start with our first of two LLM stories. Email iMessageTech, that's T -E -C -H, at M -E
-N -V -I dot O -R -G. Text or WhatsApp, 804 -442 -6975. Thank you so much for listening, and do make it a great day. We will continue in just a moment. All right, folks, here comes item three.
And this is one of two LLM stories. And when we merge the podcast, we will put a song in between these items. Email iMessageTech, that's T -E -C -H at M -E -N -V -I dot O -R -G. Thank you so much for listening to today's program. I hope that you all will be enjoying it as much as I
am when it finally gets released. Again, we're sorry for the delay. The first LLM story is LLMs and Healthcare. ChatGPT, Claude, and Google Overviews. This is probably important and this may be a little bit of a longer segment because it leads
to something that I actually am now gaining experience on. And this is chat GPT and health. And I think it's worth discussing it as part of this particular item. It will also be discussed in the next TSB, which will be recorded next Wednesday, not this
coming Wednesday the 28th, but the Wednesday afterward. API for use in healthcare, followed three days later by Anthropic's announcement of a healthcare -focused implementation of the
Claude LLM. OpenAI's stated aim is to provide specific versions of tools that are already in use in the healthcare sector. Quote, giving organizations a secure enterprise grade foundation while supporting HIPAA compliance. I'll say bull. Now, I can surely
give ChatGPT the link and say, analyze the article from this link. Titled this. And it could in theory. Give me a logical perspective. But. Granted. I'm relatively new. To this. And I'll talk about my use. As it regards to healthcare. As soon as we're done reading this. And any comments.
I may have decided to pull. Notably. Sharing medical data with ChatGPT Health removes the HIPAA protection from those records, according to Sarah Gustafson, Senior Counsel at the Electronic Privacy Information Center. The healthcare models
of ChatGPT will be trained with healthcare workflows and promise transparent citations as well as integration with institutional software policies, access controls, and user management through SAML, SSO, and SCIM, and support for HIPAA compliance.
Quote, patient data and PHI remain under an organization's control with options for data residency, audit logs, customer managed encryption keys, associate agreement BAA with OpenAI to support HIPAA -compliant
use. The final paragraph of ChatGPT's Terms of Service states, Our services are not intended for use in the diagnosis or treatment of any health condition. Anthropic characterized Claude for healthcare as HIPAA ready and says it is
set up to connect to the Centers for Medicare and Medicaid Services CMS coverage database. The International Classification of Disease 10th Revision ICD -10, and the National Provider Identifier Registry. Certain subscription plans allow Quad
secure access to patient lab results and health records. Meanwhile, Google has removed AI Overview results. From a few specific search queries after investigation by The Guardian revealed the site offering false and dangerously decontextualized
medical advice. Examples include harmful diet recommendations for pancreatic cancer, and misinformation about test results for liver function and cancer. Kim Commando mentions not to upload health data to these chatbots. I've heard that on her minutes.
If you do, try to block out any personal info tied to you as it can have that. I'll also add here that you could probably use temporary chat, which does not allow ChatGPT to remember anything.
Neely and Murray have great points here. Read what they have to say. And you can find a link in our show notes. That completes Top of the News. And we will continue with my thoughts now on what I've learned. So... On Saturday, one
day after I fell sick, I decided, as a free user, to subscribe. Not as a paid customer, but to keep everything in one place. On Friday, I told JetGPT very specific symptoms that I was having, and said, What could cause this? It gave me some
ideas, and I read them carefully. I asked what to do. It said, I think it would be best to do this, i .e., maybe rest, maybe... Take light sips of water. Maybe it's dehydration. Maybe it's, you know, various other things. It didn't say it was. It never insinuated anything. Then,
on Saturday, I thought I was getting better. But I decided I think I want an account. And we'll talk more about ChatGPT specifically on Throwback Saturday Night and our next TSB. I'm not going to read the technical aspects of the notes as I did on Saturday's Independent Artist Spotlight Show. But I can tell you that I asked
the AI to give me a podcast summary. That I could use as a technical reason to use the bot. Well, I replaced my doctor. Not permanently, mind you. But as a, this isn't a bad sign. You can treat this at home. If you have these things. And you
have this food. you do these things okay so bread food rice food you know eggs food applesauce food unspiced food like chicken turkey when i screwed up having barbecue from wingstop which
Messed me up pretty bad. It turned into a long Sunday. And I also used ChatGPT as a log tool. Mainly, if it ever became a problem, I could hand my phone to a doctor with the right chat log and say, go through that. tell it to give you a doctor's diagnosis, a doctor's write -up,
which it says it can do, and then they would know exactly what to do. And that's the technological responsibility, I think, of using a tool. I'm not replacing my doctor. I'm not saying, okay, doc, I want to come in, but I'm pooping really bad, as an example. This is an example, and I don't know if I'm going to be able to make it in, because you want me to come in, but I'm pooping
pretty bad. He may just say, you'll have to go to the hospital, and you're going to need to get treated there, and I'll come and see you. That's the worst they'd say. Maybe you could order these items from the local store, i .e. an antacid or some other medication, and try
those. I'm not going into any detail of what I had or didn't have, but I continue to log until I feel stabilized enough. Although I am feeling better as I said. But it's important to know what to disclose instead. What causes this? Or what should I eat? My stomach is still whatever. I still have these symptoms. But I'm hungry.
And that's perfectly fine. whole health history. Okay? That's the goal, right? I didn't upload my whole health history, then give it something new. I'm dealing with something in the moment. It's giving me advice of what it could be and says, if this gets worse, i .e., if... it's diarrhea,
if it gets worse and you're going more often, you start throwing up, you start feeling lightheaded, you're dizzy, you're this, you're that, then I would seek medical help. You can treat this at home by resting, eating bread, rice, applesauce, you know, things that are, you know, eggs. I'll tell you what, folks. It felt so good to eat.
I ended up eating four eggs and not have any symptoms. And the day I'm recording, I am actually feeling so much better. I had a half a meal this afternoon, a recording date, which included fish, rice, and some salad. I ate the entire fish. I still have some rice, some salad, and the pita left. But I tell you what, I feel pretty good.
I've got a hot chocolate here to help with any throat issues because there are times that I start coughing or want to clear my throat, and the warm liquid will assist with that. Or I could have some sort of water, whether bottled or canned. But the point I'm trying to make here is that it's telling me, that I made the right call. I did not force anything. I did not give it more
than it needs to know. And it was giving me simple advice. So, I think now I will actually read the document. that I titled, AI Use for Diagnosing Health Concerns from a Tech Perspective. And we will actually dive into this as part of a deeper discussion on AI later on. The tech angle.
AI as a thinking partner, not a doctor. What this experiment shows isn't that AI replaces doctors. It's that it can act as a real -time thinking partner. So I asked it, what do you think I should do? Should I go to the doctor? Is this urgent? Can I treat this at home? What
would I need? Right? You had a... mild, but uncomfortable health issue. So it's not even telling you publicly what I had, and I'm not telling you publicly what I had. I used diarrhea or stomach issue as an example in the talk. I'm not saying publicly whether that's what I had or not. As an emergency,
not as something that clearly required urgent care, just enough to make you unsure. So again, you had a mild uncomfortable health issue, not an emergency, not something that clearly required urgent care, just enough to make you unsure. That's the gray zone where people usually panic over Google. Worst case scenario or ignore symptoms
entirely. I didn't ignore those symptoms, ladies and gentlemen. I was purely saying, this is what I have. Is this urgent? Is there something I can do? What's going on here? Instead, you use the conversational AI to... Talk through symptoms in plain language. Sanity check decisions. What
to eat. What to avoid. That's in parentheses. Slow yourself down enough to make measured choices. From a tech perspective, that's important. AI lowered the cognitive load. That's important. It didn't... Make me overthink. People probably overthink and probably would overreact. Oh God, I gotta go to the doctor. I've got diarrhea.
Or I've gotta go to the doctor. I've got whatever. Sometimes you can just treat this stuff at home. Again, I'm not publishing any medical anything. Those who may know what happened to me are close. friends, and people I trust. I'm using these as an example as part of talking on a podcast. It didn't diagnose. It helped you organize your
thinking. And I wasn't saying I was ever diagnosed by AI. It knew it was potentially GI because it knows by its models that it's stomach -related because one of the symptoms was diarrhea, right? So if that was the case, then it knew it was GI? I didn't know what all that meant, but it
does. That's its job. Why that matters technologically. AI is always available. unlike doctors, friends, or hotlines. And that's important. You can't call a hotline. Some friends you trust may not be available or may not have the knowledge to know or advise you what to do. And hotlines, I don't know if I would call a hotline to get advice like that. It's non -judgmental, which
makes... people more honest. It didn't judge me. It said, this sounds like a GI problem. I would recommend rest right now. Don't eat anything. Just relax. Take it easy. Turn on Nature Space. Turn on SiriusXM Yoga or any of the other channels I told it that I listen to on a regular basis. to keep myself calm, relaxed, and in the frame
of mind as part of my everyday life these days. It can respond dynamically as conditions change. Diarrhea is gone. Burps remain. So... if the condition was that I had acid burps, but the diarrhea was gone at one point, then it could advise me, okay, if this is consistent, maybe
you should get some type of medication. And it named the type of medications to get. And it said, get one, don't get all. That's important. That loop, Observe, reflect, adjust is basically a feedback system. You were running a human -in -the -loop health model. The subtle win. The
healing didn't come from advice alone. It came from thoughtful planning. Okay? So, eating small portions. avoiding known irritants, soda. So I bought a soda one day, but I didn't touch it. And even when I did touch it, I took slow, deliberate sips, put it down, and left it there. Monitoring changes instead of reacting emotionally. So for
example, okay, it's 10 o 'clock at night. I'm done with the diarrhea. I still have this issue or that issue. What's next? Try and get some sleep. See how you feel in the morning. If you have anything overnight, let me know and we'll reevaluate. If it seems like you need to go get medical care, I'll advise that. Otherwise, I'm here for you. A, I didn't replace intuition.
Amplified it. The responsible take. AI works best in low -risk situations as a support tool, not as a replacement for medical care. And I think that's the very important thing. If it said, this sounds like urgent care, you need to do this, then that's exactly what I would do. I hope it never comes to that. But now I
have the tools to assist me, i .e., if my symptoms were. I'm coughing. I can't breathe. This feels to me like an upper respiratory infection. I get them regularly. I've tried to rest, but I can't rest. I can't sleep. When I cough, it hurts. That type of thing. Then it would probably say, go get medical care. It's great for reassurance,
decision scaffolding, knowing when to wait versus when to escalate. And that's a very defensible, forward -thinking tech use case. So that's that file. And that, I think, is important. Because if we take nothing else out of this, then what would AI be used for? I did use it for finding
out when an artist is an independent. I've used it more recently. and discussed vending machines and other accessibility stuff. I even had it generate an email which I submitted because there was a serious issue in the ChatGPT app itself
as I was learning it. So it can be very useful in the right situations. You've just got to be smart and not share every little detail. Or maybe, if that's what you're going to do, switch off the item to let it learn from your account so that it can be trained based on everything you
have. cause problems and i'm not going to say hallucinate because we'll talk about that more on tsb there's a lot i want to get into when it comes to the aspect of this whole ai thing because the hallucination topic was very fascinating when i was talking to it about that specifically
Now, it could say, I don't hallucinate, I make up stuff. Well, maybe in the tech press's mind, that's hallucination. And that I understand. But that's because it's just fed a lot. And we'll discuss all of that later. But in the context of whether I would use it in, like, a hospital as part of clinician work? Maybe in the right
circumstances, i .e., a chat for each client, the account not being in the I'll share everything so that the models learn from what I'm doing option. I forget exactly what it's called at the moment, but we'll dive into all of that when we do our AI chat. Suffice it to say, I've got a lot of ideas coming for that podcast, and I'll
start work on that very soon. In the meantime, as I said, I wanted to include this as part of this discussion so that you are aware of how I used it. And yes, I did give a couple of symptoms, but I'm not giving every little detail because it honestly doesn't matter. I'm good now. It may take me another day to get this particular podcast out because I had several days off and
I'm doing this mid -afternoon and I don't know if I want to stay and do this all night. In the meantime, I do want to play the track which will be coming up next. I'm not going to tell you what it is do this on the fly and I don't know what I'm going to pick. We will continue with the rest of the news right after this track. This is Sands episode 38 covering the newsletter
for January 13th, 2026. back to sans 38 and you're
probably saying you should have released this a long time ago yes this is absolutely true but after i started recording this program all hell broke loose i got sick I got better. The computer
actually had technical problems that needed resolving. I got sick again. The 40 cell display I normally use ended up having dot issues. Where now it's out for repair. My. 32 cell braille sense. Is
having issues. And then I had to. Bite the bullet. And get the e -reader from humanware. So there's been lots of technical debt. So that's really been. the real reason why this podcast has been delayed. But I'm hoping to get back into things. Number one, the orbit arrived to get repaired.
I have an e -reader that I can use. And between everything, I can, in theory, now try and get back into things. And everything is going to be... All well and good. So we're going to start the rest of the week's news for this newsletter that we started covering way back in January.
Item number one in this segment is LLM APIs targeted by threat actors and gray hat hackers. And this could still be going on today in different situations. So even though this podcast is going to be out
In April, because of all of the issues that we've had going on, some of this news that you're going to be hearing may still apply. So, as I say, better late than never, because I tried to get this out before the computer actually died. So, I at least tried. Threat research from a grain
noise honeypot shows two recent campaigns probing the security of LLM APIs by way of misconfigured proxy servers. The first campaign spanned from an October 2025 To January 2026. Exploiting server
-side. Request forgery. Vulnerabilities. To target. L. L. A. M. A. Or. Actually this is. Olama. Sorry. So to target. Olama. Model. Poll functionality. And Twilio. SMS webhook integrations. The attackers
used Project Discovery's OAST out -of -band application security testing infrastructure to confirm successful SSRF exploitation, leading Gray Noise to believe
the attackers were likely researchers or bug bounty hunters. The second campaign, however, was most likely a professional threat actor conducting reconnaissance by methodically probing at least
73 LLM model endpoints over 11 days starting December 28, 2025. Again, I want to just make sure that while the podcast is going to be released well after this was published, and the link is going to be in the show notes, this type of thing
could possibly have gone from research to a possible problem. Now, I haven't heard of this happening yet, and because we're so late, I want to make it known that this could, in theory, be happening. Yes, I did that on purpose. The attackers generated
over 80 ,000... in that time, hunting for misconfigured proxy servers that might leak access to commercial APIs. Every major model family appeared in the probe list. Data associates the attacker's IPs
with extensive previous CVE exploitation, and the researchers posit that the threat actor is building target lists as part, of a larger pipeline. Groenoise recommends users configure Ollama to
only accept models from trusted registries. Set up alerts for rapid -fire requests and fingerprinting queries. Block OAST at DNS, rate limit suspicious ASNs, and monitor jabbled fingerprints. The blog
post contains network fingerprints, OAST callback domains, and IP addresses to block as part of defending LLM infrastructure. Dukes writes, API security continues to be a concern and misconfigurations are the leading cause. Software developers should
reference the OWASP API security project's top 10 list as a guide when building the API. And while you're at it, fund the OWASP nonprofit. They do great work. Neely, who was first, writes,
the action here is to make sure your threat hunters Are incorporating. IOC's. And. That you're taking steps. To protect. Your LLMs. Including. Gray
noises. Suggestions. Of. only allowing models from trusted repositories watching for enumeration patterns and rate -limiting blocking suspicious networks and domains. And all I'm going to say is this. While I understand this is late, and
later than I really wanted, because we really tried to get Sans Newsbytes out within one week of it hitting my inbox, and all the trouble that I discussed in my introduction to this segment, we've really got to still be aware of this research
in case you've noticed that things have changed. It's not necessarily about the fact that it's late news. Well, Jared, you shouldn't even be covering this because it's months old. Yes, you're absolutely right. But the problem is that I started recording this when this was actually first put
out as a newsletter. So now all I'm going to do is finish the work that should have been done in the first place if it wasn't all of the issues I was trying to deal with that I stated above. And part of that delay was getting sick. And I'll tell you what. What I was going through wasn't all that fun. It wasn't necessarily COVID or a cold or anything. It could have been a stomach
bug or two because I had that happen to me twice in the matter of several months since I first started recording this program. All we're saying here, folks, is that This is something that still should be at least read. Understand it, because once this really gets out in the public, while
we may not know about it now, it could be part of some arsenal later on. And I think that's really one of the reasons why I still want to cover this because, yes, it's late now. I understand
that. But could this become a future reality problem? Hmm. Something to think about. Instagram is in the news, and we always know that Instagram has always hit some kind of news. So, since we're still in January for this newsletter, it wouldn't
surprise me if other Instagram stuff has come out, but we've got something interesting coming up next. As we're going to talk about password reset emails. And I think this is still appropriate. Again, even though this is months afterward, which was never my intent. But we always hear about issues where password reset emails... could
trigger things that you're not expecting. I'm Jared Reimer. You've got Sans 38. We will continue in just a moment. Let's talk a little bit about
passwords. In this story from the Sans News Bites that we have been delayed on, this one's titled, Instagram password reset emails are unrelated to alleged... data breach so again this is going
to be covered late because we want to finish this before we move into other sans content and i feel that that's more appropriate because we've had this here for a while So, let's do this by
reading it. On January 9th, 2026, Malwarebytes warned that cybercriminals stole the sensitive information of 17 .5 million Instagram accounts, including usernames, physical addresses, phone
numbers, email addresses, and more, and posted a screenshot showing an Instagram password reset email. However, the data in question and recent reports of unsolicited Instagram password reset
requests are not related despite surfacing online simultaneously. Instagram announced via social media on January 10th that while there was no breach of their systems, the company has now
fixed A vulnerability that had allowed a third party to trigger password reset emails and directs users to disregard the emails. Well, even as late as this is, that's always good news. Well, we did have an issue, but it's unrelated to what...
was actually going on. So, that's always great news. And as late as this is, I do want to say that overall, this is great news, right? There wasn't a known breach that claimed to reset emails, but a similar issue that could be unrelated. And I'm not saying back then that there was an
issue with Instagram. Remember, it's a Facebook product. No recent data breach has been confirmed by Meta, but Malwarebytes was likely referring to a dataset recently published in a hacking forum and listed on Troy Hunt's site, Have I
Been Pwned? And that is still a great site, and TSB has not been really keeping up with Have I Been Pwned in recent times. Not because we don't want to, but because there's so much happening. And I think people just need to find time to
really check these things out on their own instead of me just saying, okay, this is what they're reporting. Because honestly, we don't see a lot of stories like they do. And... Everything else.
So... That is... Where we are. So in parentheses, H -I -B -P. So if you haven't been there, have I been? P -W -N -E -D dot com. Find a link at
emailhostsecurity .com under our resources. Our resources can be gotten to directly at emailhostsecurity .net. H -I -B -P is one of two websites that we actually recommend. The other is ExposedOrNot, X -P -O -S -E -D -OrNot .com, which you can also
find on our website. The data allegedly contain user information obtained via API scraping. Some cybersecurity researchers speculate that the breach happened in 2022, but the data's provenance, authenticity, and age have not been publicly
corroborated. bleeping computer posits that the data may be a compilation of previously scraped information about multiple sources over several years. And we know by default that There could
be a breach, okay? But actors may not use anything from it. But then, there could be another breach. Let's say you're not in the original, right? So, if Instagram was breached, you don't have an Instagram account. But all of a sudden, your information comes out because some other breach
happened. But yet, it's tied, for example, to a story like this one, where you could get an email saying that your account was compromised, yet you have no Instagram account. So they could package a dataset, call it an Instagram breach, even though Instagram, in its right, could say,
no, that's not our breach. So, that's that. I did see the Malwarebytes story in their newsletter.
A combination of things is what I figure, and I did not post the article. So, back then, I decided not to publish it, because it seemed to me That there was a whole combination of things going on here. And this could happen even now.
We covered this in our last TSB program in passing. So if you listen to TSB in January. You will hear us talk about this particular story because Nick mentioned it, and yes, I still think this is going on today. Neely writes, It's a good
time to check the HIBP website for all your email addresses, as well as make sure that your password practices are up to snuff. You know the drill. Use good passwords, don't reuse them, and enable MFA and passkeys for, I'm sorry, wherever supported. If you have any doubts about the security of
a password for a service, update it using their password changing mechanism. Disabled accounts for services you are no longer using. I'm sorry, it says disabled, not disabled. And I think that is a great idea. if the service allows you to do that. Keep a record of that action. Sometimes
it takes a bit to close an account. You can also check exposed or not whose link is on our website. at emailhostsecurity .net. The short link to the resources page, as I talked about in this segment. Let's continue now with Breach Forums.
And even though this is late, I always love stories like this because we know Breach Forums was never meant to be good. And if we can get a couple of stories, even though this is going to be months late, I win completely. Right, so... Item number
three in the rest of the news says, breach forum member data leaked. ReSecurity has analyzed a database of 323 ,986 forum members' records alleged to identify... administrators, moderators, and users of the latest incarnation of a cybercrime
forum called Breach Forums. Let me make this very clear that at this time, I do not know if they are still around or not. Any incarnation of this forum, not good the company's threat intelligence team believes the database and associated
leaked data contain information that may be useful to law enforcement pursuing cyber criminals some of the records Identified. In the database. Are definitely. Authentic. And can. Be. Cross checked. With other sources. Regarding specific. Actors.
However. Some records. Have been edited. Removed or contain. Non. For example, replaced on IP 127 .0 .0 .9, which is likely an OPSEC measure taken by the actors administering it. The last
registration date in the newly leaked user database is from August 11th, 2025, which is the same day the previous breachforums .hn was closed. Hmm. Race security contrasts this breach with
previous examples of information released by threat actors asserting that the events Involving the compromised. Breach forums database. Are. Different. From this activity. And contain. The meta data. Of many. Notable bad. Actors. So.
This. Is a taste. Of. Their own. They think they're untouchable. But are they? I don't know. And
the fact that we haven't seen anything that... Terry Ring, Preston Gaylor, Nicholas Jackson, or others that might end up sending me things about this type of thing. They haven't seen anything. Hmm. I wonder. I write, this is always good news in my book. If we can catch some. We can slow
down the services, or have them stopped too. Neely, the face of the newsletter it seems, writes, It seems the tables have turned. In this case, the breach forum is itself breached. That said, releasing the database is not without risks. The database was released by James. And Resecurity
hints they know who that is. If they can figure it out, so can others. If you want to analyze the data, obtain a clean copy from Resecurity. as there are other copies which contain malware yes mr neely the tables have turned and this
i think we can agree is good Here comes our next item. California Privacy Protection Agency finds
Texas firm for failing to register as a data broker with the state. California's Privacy Protection Agency is finding Rickenbacker Data, DBA, Data
Masters for failing to register as a data broker in the state of California. The decision asserts that Data Masters bought, repackaged, and resold contact data of people with a variety of medical conditions, so the information could be used for targeted advertising. The action from CPPA
finds Datamasters $45 ,000 and orders the company to stop selling data belonging to California residents. Data Masters was also ordered to delete all Californians' personal information it holds by the end of December 2025. In a separate action,
see PPA, find S. NP Global $62 ,000 for failing to register with the state as a data broker. The issue was due to an administrative error. Neely writes, An indicator That's consequences
for failing to follow CCPA are real. Double check the applicability of CCPA to your datasets. Datamasters claimed they were exempt from CCPA because they didn't operate. California but it's the processing
obtaining of data that belongs to Californians without deleting it within 24 hours which brings CCPA into play if you are a data broker for the California, if you are a data broker for Californian data, make sure that you register. The California
Delete Act expects more enforcement as single point opt out is implemented in the new drop. platform dukes start this whole thing off with californians consumer privacy act ccpa is pretty
specific regarding data brokers one of the main requirements is To register within the state. And BTW. That's an annual requirement. Looks like California is getting serious about getting out. I'm sorry. Of ferreting out data brokers.
operating illegally in the state. Murray also has a great point. Given how lucrative the data broker business is, these fines are not likely to be effective. The orders may B. Kudos to California,
its law, and its agency. What do you say? Well, I haven't seen any more about this, at least in public circles. whether in podcasts or even brought to my attention. But companies should be ashamed of themselves for collecting data
because that's what they do. They collect it. They sell it. And frankly, we don't even know who the they are. I'm sorry, but when I initially read this section of the newsletter, I was happy to see that they got fined. Yes, it's probably
a small drop in the bucket. And they'll just pay them and say, go away. They may delete the data. Then they'll go back and collect it until they get caught again. But the fact that they got caught and they had to pay makes them slow down. Whether it stops the whole thing, I don't
know. I'm not ready to go that far. But, I still want to hear what you have to say. Send me an email or iMessage at tech, that's T -E -C -H, at M as in Mike, E -N as in November, V as in Victor, I as in India, dot org. Text or WhatsApp 804 -442 -6975. And of course, you can leave
your comment on the phone line at 888 -405 -7524. Because this is going to get very interesting.
I assume that this is going to be probably one of the biggest stories of the year. as we'll be approaching mid -year soon. And we haven't seen anything new, but yet again, I haven't been reading sans on a regular due to the aforementioned issues I've told you about in segment four of
this podcast, which is why it's been delayed so long. Suffice it to say, it wouldn't surprise me if other SANS news bites that we have not been able to cover and will not cover because of how far we're behind has covered this. Starting with the next podcast of SANS, we're going to just move forward instead of trying to catch
up completely. Please understand that we're trying the best we can. And with everything going on in the various podcasts we have, there's only so much we can do. So what do you say about this? Yay? Nay? I don't care? F you? Go ahead. You
can say F you. I try not to make this a swearing program. At least on sands I don't. I try and make it my own without getting my emotions caught up in everything. I try. But you tell me. Yay, nay, what? Spanish authorities arrest 34 in connection
with Cybercrime Network. And again, we are very, very happy to see things like this because we don't see a lot of it. And I know one of the things that I want to see more of is news just like this. We always talk about ransomware, you know, and all of the bad things. And we don't
see good things. Now, will it lead to any time? I don't know. But the fact that people are arrested and charged with things slows things down. Is that a good thing? Overall, I believe it is. The write -up for this says, Authorities in Spain have arrested 34 individuals in connection with
cyber fraud conducted by an international criminal group. According to investigators, the group is responsible for fraud losses of more than five Kaba 93 million euros. I believe that's
a comma. I don't know what exactly this is. It may be 5 .93 million. I'm not 100 % sure. Because I see a dot. It's not a period in either braille code. So, I'm not exactly sure. But anyway, let's just say it's $5 .93 million, which is quite
high. Law enforcement recovered a small portion of that amount by freezing bank accounts and seizing cash. Well, seizing cash is probably easy. Bank accounts, I don't know how law enforcement would seize that. I would assume that would be
up to the bank. Cash, you could take and say, that's not yours. Not until we prove you earned it. Business email compromise attacks. Illegal vehicle trafficking through shell companies and other fraudulent activities. The law enforcement
action was conducted by the Spanish National Police. national in cooperation with the Bavarian state criminal police office and this word is bear chess and the next word is Bendis criminal
it's a criminal lemt so or no it's landis criminal lemt uh so it's gonna be b -a -y -e -r -i -s -c -h -e -s and then l -a -n -d -e -s -k -r -i -m -i -n -a -l and then AMT and support from
Europol. Europol writes that the cross -border collaboration between Germany and Spain included Spanish investigators receiving analytical support, the exchange of intelligence, and the deployment of two German officers on site during the action day. Europol supported the action through a range
of services including information analysis, a data sprint held in Madrid, and on -the -spot support. Neely writes, it is nice to see cross -border cooperation continue to shut down criminal operations. This group has a global presence
operating in Nigeria and abroad. Spread over about 60 zones with 200 members each. and has a total membership of about 30 ,000. They were known for recruiting money mules from impoverished areas with high unemployment rates. I agree,
I write. I love great news like this. Keep it coming. And yes, Money Mules. That was actually talked about in recent episodes of a podcast that deals with crime. More specifically, different
types of crime. But they do talk about bank crime and what's happening. And the podcast, in theory, is mainly meant for people who don't understand a lot of what's happening recently. This podcast. was talking about business email compromise, and they made it known that they didn't know
what it was. And while I can't necessarily blame the hosts per se, when a host brings something up, it's a good idea if you don't know what something is, to just indicate that you don't know what it is. It's okay to read and give your thoughts about what you're about to talk about, and then
say, we want to look this up, we don't really know what this is. But to come out and just say, we don't know what this is, but we're going to talk about it anyway, makes it look bad, in my opinion. Now, I'm not going to sit here and bash podcasts. That's not who I am. We here at the Jared Reimer Network want all podcasts to succeed, and we definitely think that there is an important
need for a podcast that is simple, talks about various types of scams and crime, It's pretty much what scams are. It's a form of crime. The podcast itself is titled Scam Squad. It's ran by two individuals. And, you know, if you find it of value, go ahead and check it out. I'm not going to say publicly that they should go because
people... probably would say that this podcast set needs to go. And that's not what I'm about to do. I'm not going to do that publicly. You know, what we might say privately could stay that way. But I really think that education is key. And each podcast is going to do it differently. So, let's just leave it at that. And we'll go
from there. If you've heard of the podcast and you find it of value, let me know. Because I want to make sure that people are educated the best way that they can be. When we will continue with the podcast, Or better yet, when we continue with this podcast, we're going to talk about a printing error. And this printing error has
got to probably be the worst I've ever seen. So let's see what this is. I'm Jared Reimer. This is SANS 38. Printing error prompts recall
of nearly 13 ,000 recent Irish passports. Talk about a problem much? A software update appears to have caused Irish passports issued between December 23rd 2025 and January 6th, 2026 to be
printed incorrectly, rendering them invalid. Ireland's Passport Service says that in order to mitigate against any possible travel issues, they have notified border authorities worldwide through the International Civil Aviation Organization,
ICAO, as well as Irish Border Management. The documents in question are missing the letters IRL. which means they are not compliant with border control and e -gates requirements. The
Passport Service has contacted all 12 ,900... for affected customers asking them to return their incorrectly printed passport books and cards, informing them that they will be issued
new documents with new numbers. What a costly mistake! For that many, there will be some Upset people, and that's putting it mildly. Neely writes, Use caution with move fast and break things and have a plan for rolling back or fixing what you
break. In this case, physical replacement of passports is a bit... higher impact than a case where you may have had to update data and possibly send a notification. When things do go sideways, make sure that someone has the customers back.
In this case, the Irish Passport Service has been emailing affected customers, updating their website, and setting up a dedicated customer service team for those traveling immediately as well as covering any costs of reissued passports or visas. The good news is that the customer,
by the way this reads, doesn't have to pay a dime. I guess that is a good thing. I could see them being outraged if they did. What a costly printing mistake, even months later and reading it now. I don't think I've seen anything as bad
as this, even though I've not read Sans since on a regular. But I know that my sources haven't reported anything, but that definitely doesn't mean anything. If you've seen something, even if it's late, I'd like to at least review it.
I believe this is our last item of this program and we can finally get old news out the door. Again, I completely apologize for the fact that this took so long. And now that I think we're stabilized, we'll be able to do more of these on a regular basis. So, let's cover our illustrious government now, because this is a CISA item,
and we know that CISA itself has been under fire with funding and other things, and remember that this is the company that says, quote, you need to do, end quote, but, quote, we don't have to. End quote. Meaning, do as I say, not as I do. This item says CISA retires 10 emergency directives.
And this is probably a good thing because if they're not needed, then that is the right thing to do. But again, this is the same company who is tasked by the government to keep everyone as secure as possible, yet they themselves, meaning the government as a whole, I'm not saying CISA doesn't, I'm saying the government doesn't. And
to be honest, I'm not attacking just the U .S. government, I'm attacking government in general when I say that they don't update. On Thursday, January 8th, 2026, the U .S. Cybersecurity and Infrastructure Security Agency, CISA, retired
10 emergency directives issued between 2019 and 2024. In a press release, CISA writes that a comprehensive review of all active directives determined that required actions have been successfully implemented or are now encompassed through binding
operational Directive BOD 2201, reducing the significant risk of known exploited vulnerabilities. The retired emergency directives include one directive from 2019 instructing FCEBs to mitigate
DNS infrastructure, tampering three directives from 2020 regarding a variety of Windows vulnerabilities, four directives from 2021 including the SolarWinds Orion code compromise, Pulse Connect, Microsoft Exchange On -Premises, and Windows Print Spooler.
One from 2022 regarding VMware vulnerabilities, and one from 2024 regarding the nation -state compromise of Microsoft corporate email systems. I write, I was almost going to write that this was a bad idea, but reading Neely's response
seems quite fair to me. He writes, the emergencies have passed. Countermeasures and updates are in place for those specific flaws. The focus needs to be on keeping things secure slash updated. CISA slash DHS's continuous diagnostics and mitigation
program is designed to do this across the federal government. It boils down... to keeping an eye on the ball and identifying and addressing issues before a directive is issued to do so, ideally with a dashboard for management to review in lieu of another meeting or report. Or and report.
Dukes also writes, the key word is action. emergency, end quotes. And yes, directives issued in the last six years should have been implemented in that time frame. The bulk of the now retired directives were issued to mandate patching of specific vulnerabilities that have been around for a year or more. I'd like to believe they
were all patched. So would I, guys. So would I. And I guess we're just going to have to see what happens overall. The editorial board of SANS Brian Honan. Kurt Dukes. Chris LG. David
Holzer. Ed Skoudis. Gal Spitzner. Jake Williams. Dr. Johannes Ulrich. John Peskitor, Josh Wright, Kathy Bradford, Lance Spitzner, Lee Neely, Mark Weatherford, Moses Frost, Suzanne Voltronaut,
William Hugh Murray, and Sands Institute is located at 11200 Rockview Pike, Suite 200, North Bethesda, Maryland, 20852. Email support at sans .org or call 301 -654 -7267 for assistance. Well, ladies
and gentlemen, I am so happy that you were able to trust us in trying to get the news out because we definitely want to get that news to you. And we can't do it alone. Unfortunately, do... To
other. Things. We were not able to do this. As fast as we normally do. And. We. Definitely. Want. The. Particulars. Of. Making sure we're accurate. and doing everything we can. Speaking
of which, I have enabled the ability for transcripts, and we're going to try this again. And starting with our next podcast of A Day in the Life, we are also enabling the transcripts. And hopefully that might help some people because accessibility is something that I want to try and do. And since
it's available to me, we're putting it on the highest quality possible. And hopefully it is good enough. The transcripts may not be fully accurate, but it should be accurate enough. So, we'll just have to see how it goes. And yes, if we end up playing music that's vocal on the security box, it'll more than likely transcribe
the lyrics of the tracks. I've seen it. It's quite interesting. But you could ignore that if you want. I don't know which programs support it, but we are going to try it anyway and see how it goes. So, we'll catch up with you all next SANS episode, and we are going to go ahead
and just move forward and go to the most recent episode available. Catch up with you all next time.
Welcome to podcast 280 Intermediate D. I'm Jared and the podcast is now available for everyone. On this podcast, the Evolution of Cyber Security, part 3: Organized cybercrime and social engineering, part 5. The podcast...
Hello gang, welcome to the security box, program 280 Intermediate E. On this program, we're covering something called PHP2Shell two vulnerabilities that were disclosed the same day. Not only that, we have a third vuln...
Welcome to Sans, episode 43 where your blood may decide to boil. I'm Jared, and this episode will cover the Newsletter that was released on October 13, 2026. On this newsletter, there will be some good news items that...
Welcome to podcast 280 Intermediate D. I'm Jared and the podcast is now available for everyone. On this podcast, the Evolution of Cyber Security, part 3: Organized cybercrime and social engineering, part 5. The podcast...
Welcome to Sans, Episode 42. Its been awhile, and this episode explains why. I hope that we can continue Sans, I'm going to start on the next issue very soon now. If you want to read the newsletter, please use this link...