Podcast Summary: Talkin' About [Infosec] News, Powered by Black Hills Information Security

Episode: Chinese agent tried to recruit Stanford Student - 2025-09-02
Date: September 6, 2025
Hosts: John Strand, Bronwyn, Mary Ellen, Wade

Overview

This episode centers on recent cybersecurity incidents and trends, including software breaches, vulnerabilities, the questionable ethics of residential proxy networks, the emergence of AI in malware and security defense, and an intriguing story about a Chinese agent trying to recruit a Stanford student. The Black Hills Information Security team discusses the technical details of these stories, offers professional and often irreverent commentary, and shares their personal takes on infosec developments and the current AI bubble.

Key Discussions & Insights

1. SalesLoft Drift Integration Compromise

• [03:00-05:29]
• A notable breach with the SalesLoft Drift integration caught the attention of Google Threat Intelligence Group (TIG), which posted an advisory.
  • SalesLoft: Not the same as Salesforce, but an 11-year-old sales conversion company with a Drift platform used on customer sites.
  • Compromise Details: Salesforce stored keys for SalesLoft Drift, which were then accessed by attackers after the Salesforce instance itself was compromised.
  • Advice: “If you use Sales Loft Drift, you need to treat all authentication tokens stored in or connected to the Drift platform as potentially compromised.” — John Strand [04:20]
  • Companies Affected: Google (initial disclosure), Zscaler, Palo Alto Networks, possibly others.
  • Mitigation: Reset credentials and tokens if you use this integration; SalesLoft and Salesforce have pushed updates and advisories.

2. API Integration Security Risks

• [05:30-07:00]
• The team reflects on how interconnecting business applications (often for convenience) increases security challenges.
  • “The API integration world is not always done in the most secure way...a lot of it is based on convenience and not so much on security.” — John Strand [05:51]
  • The hosts note the recurring issue of insecure authentication and data transport.

3. Legal Botnets & Residential Proxy Networks

• [07:00-13:40]
• Reddit Story: A user admitted to hosting DSL Route proxy devices at home for payment ($250/month). DSL Route, registered in the Bahamas, is used to gain access to reputable US-based IP addresses for questionable purposes.
  • Context: Hosting proxies is attractive for threat actors as US IPs typically have a better reputation and face fewer blocks.
  • Ethics and Risk: “If you have to ask Reddit, it sounds like a here hold my beer moment.” — Bronwyn [13:33]
  • Legit Alternatives: There are above-board proxy providers with Know Your Customer (KYC) requirements (Bright Data, Oxylabs), but sketchy actors abound.

4. Velociraptor: Living off the Land Attacks

• [13:40-17:06]
• Threat Intelligence: Sophos reported threat actors are abusing Velociraptor (a legitimate IR tool) to install VS Code and create C2 connections, increasing the stealth of attacks.
  • “Living off the land is definitely like the current hotness as far as attackers are concerned.” — John Strand [15:54]
  • The team recommends defenders set up alerts for unusual tools (like VS Code) on endpoints, watch for domains like workers.dev.

5. Ransomware Updates: Sweden & Nevada

• [17:06-20:30]
• Swedish Supplier Attack: Miljo Data, a Swedish IT supplier, hit for 1.5 bitcoins.
• Nevada Government Attack: Major US state hit hard. Many services (DMV, Health & Human Services) affected, with multi-week disruptions.
  • “This one is rare for being kind of the most integrated...so many different services were taken down.” — John Strand [19:16]

6. Citrix Vulnerabilities

• [20:30-23:05]
• Citrix Bleed: Kevin Beaumont (doublepulsar) disclosed new CVEs (e.g., CVE-2025-6543), leading to remote code execution and denial of service in NetScaler and Gateway products.
  • “Patch your Citrix products.” — Mary Ellen & John Strand [23:03]

7. AI in Security: Bubble, Breaches, and Trends

• [23:05-30:00, 40:00+]
• Bullet Points:
  • AI-driven tools being found unprotected on the Internet; Cisco Talos says 1,100 Ollama servers exposed on Shodan, most dormant.
  • First AI-Powered Malware?: ESET claims to have found "Prompt Lock," an experimental ransomware tool using open-source LLMs to generate malicious code. Team is skeptical about practical in-the-wild impact.
  • “It makes sense because one of the things that LLMs are really good at is generating slightly different stuff at scale. So that’s one of the reasons why it’s so good for phishing campaigns.” — Bronwyn [28:32]
  • AI Bubble: Majority of companies that went all-in on AI now backpedaling. Terms of service and privacy policies get rapid-fire updates, indicating industry uncertainty.
  • “I think it mostly overhyped...it’s not gonna have as big of an impact as everyone thinks it is.” — John Strand [43:45]
  • Some defenders (Wade) strongly endorse AI for analyzing logs, writing IR reports, and boosting blue team productivity.
    • “It has been a game changer...summarize investigations...write me an incident report...saving me hours a day.” — Wade [45:06]
  • Discussion of smaller, specialized models as the next step after the bubble burst.

8. Salesforce and Supply Chain Breaches

• [32:00-33:55]
• TransUnion: Another data breach (via Salesforce integration). Disclosed data includes unredacted Social Security numbers, contact info, DOBs.
  • “Are there any Social Security numbers that have not been breached at this point?” — Bronwyn [33:05]
  • Also affects Farmers Insurance, Palo Alto Networks.

9. Chicken Communication (Light Interlude)

• [34:16-39:29]
• Ongoing humorous side-thread about a Canadian university’s “AI chicken translator app”, local efforts to test it with real chickens, skepticism about its value (“How is that AI powered at all? That just sounds like we’re the AI.” — John Strand [39:17]).
• Also: Discovery of McDonald's reward system API bug—enabled free chicken nuggets via password/reset exploit. No bug bounty program in place.
  • “If you’re a bug bounty hunter, I think going hacking for free nuggets is a pretty good target.” — John Strand [36:47]

10. Chinese Spy Recruitment of Stanford Student

• [51:26-55:40]
• Key story from the episode’s title: A Stanford student, taking Chinese, is approached via Instagram by someone who turns out to be a Chinese intelligence agent trying to recruit her.
  • The recruiter attempted to move the conversation off Instagram to WeChat and offered to bring her to China.
  • The student was initially hesitant to report but eventually did, raising worries about risks to herself if she returns to China.
  • Motivation: Likely seeking research insights or connections from Stanford’s AI and tech communities.
  • “He tried to get off Instagram and onto like WeChat over and over. Which that’s typical for like a move, move over to a different chat program that they own, which that’s not a red [flag]...” — Wade [54:01]
  • The team reflects on the frequency and subtlety of such state-sponsored attempts, and how “even if the hit rate is 1%”, it’s high enough for these efforts to persist.

Notable Quotes & Memorable Moments

• On security basics:
  • “It’s amazing to me how the, the more cool tools we have, ultimately the security aspect boils down to the exact same thing over and over again: secure authentication, secure transportation of data.” — Bronwyn [06:37]
• On dubious proxy hosting:
  • “If you have to post to Reddit and be like, hey, Reddit, am I… is this a terrible idea? Should I not do this? Like, it’s pretty good, pretty safe to say that it probably isn’t a great idea.” — John Strand [13:39]
• On AI’s place in infosec:
  • “AI is not going to replace anyone. It’s going to replace people who don’t use AI.” — John Strand [49:03]
• On the chicken app:
  • “How is that AI powered at all? That just sounds like we’re the AI.” — John Strand [39:17]
• On social engineering threats:
  • "He tried to get off Instagram and onto like WeChat over and over. Which that’s typical for like a move, move over to a different chat program that they own." — Wade [54:01]
• On burnout and escapism:
  • “Living a dystopian nightmare. Can I please wake up? … All you have to do is don't pay attention. Works great.” — Bronwyn & John Strand [50:46, 50:49]

Timestamps for Key Segments

• SalesLoft Drift Breach: [03:00-06:30]
• API Security Challenges: [05:51-06:59]
• Reddit Proxy/Botnet Discussion: [07:00-13:40]
• Velociraptor Living-off-the-land Technique: [13:40-17:06]
• Ransomware Corner (Sweden, Nevada): [17:06-20:45]
• Citrix Bleed Vulnerabilities: [20:45-23:05]
• AI Servers on the Internet / Prompt Lock Malware: [24:55-30:00]
• TransUnion & Salesforce Data Breach: [32:13-34:00]
• Chicken Communication/AIs in Birding: [34:16-39:29]
• Chinese Spy Recruits Stanford Student: [51:26-55:40]

Tone & Style

The hosts maintain a lively, irreverent, and conversational tone—mixing technical expertise with sarcasm, cultural references, and humor. The banter includes “ranty” exasperation at industry trends, wisecracks about product names ("Salescast," "Clever girl" for Velociraptor), and tangential story-swapping that keeps the discussion engaging.

In Summary

This episode offers a blend of timely infosec news, industry analysis, and inside perspective with a humorous twist. Listeners get up-to-the-minute details on real breaches and vulnerabilities (SalesLoft, Citrix, TransUnion), ethical debates around legal vs. illegal botnets, emerging attack methods with legitimate tools, skeptical hot takes on the AI bubble, and a chilling real-world example of nation-state spying on US university students. All this is rounded off with quirky stories about chickens, free nuggets, and the personal applications (and misapplications) of technology.