Alice in Supply Chains

Tenchi Security

In 2022, Tenchi Security created the Alice in Supply Chains newsletter to share and highlight stories on third-party and supply chain risks and attacks, their impacts on services and businesses, and how the industry is moving forward to manage these risks. Following the meteoric success of the newsletter, we've created a podcast! Every month, hosts Adrian Sanabria and Alexandre Sieira will discuss the top six stories from each monthly newsletter, and their thoughts on the future of third party cyber risk management.

  1. 1d ago

    Episode 23 | November, 2026

    In the November 2026 episode of Alice in Supply Chains (episode 23!), Adrian and Alexandre start with recaps of Tenchi Conference and Alexandre’s Third Party Risk Association panel on TPRM risk domains, then dig into this month’s stories.Story 1: Third-Party Cyber Risk Management in Brazil - 3rd Edition - 2026: https://www.tenchisecurity.com/en/2026-tpcrm-reportTenchi’s 3rd annual TPCRM report finds 76% of companies hit by an incident in the past 24 months traced at least one back to a third party, and 84% rank near-real-time monitoring as a top success factor — yet self-assessment questionnaires remain the most common method. Alexandre argues the real predictor of a mature program is risk governance that keeps the business from simply overruling security. Survey participation also funded donations to GRAACC, a Brazilian children’s cancer hospital.Story 2: CISOs Have Spoken: Risk Reduction, Not Another Cyber Risk Rating: https://www.forrester.com/blogs/cisos-have-spoken-risk-reduction-not-another-cyber-risk-rating/Forrester’s Paul McKay says cyber ratings have had their day. Adrian questions whether they ever worked (and plugs Kayne McGladrey’s upcoming book, Cyber Risk is a Myth), and asks why ratings vendors don’t publish how often their scores predicted actual breaches. Alexandre notes Gartner and IDC are reaching the same conclusion: outside-in scanning sees a tiny slice of a vendor’s environment. In Tenchi’s data, most companies score A or B from the outside, but scores fall into a bell curve once inside-out data is added. Plus: SBOMs are collected but rarely used.Story 3: Field Auditors Flee Firms, Favoring In-House Functions: https://www.linkedin.com/posts/petestrouse_ive-had-5-messages-from-external-auditors-share-7519195396804395009-hdv7/Flat pay, heavy workloads, and clients diverting budget to AI tokens are squeezing external auditors, which matters for anyone relying on their reports. The hosts discuss the pyramid model of audit and consulting firms, Jevons paradox, and why AI replaces tasks, not judgment.Other Links: Hospital do GRAACC: https://graacc.org.br/Kayne McGladrey’s book preorder: https://www.routledge.com/Cyber-Risk-is-a-Myth-A-Business-Approach-to-Integrated-Risk-Management/McGladrey/p/book/9781041249054

  2. 4d ago

    Episode 22 | October, 2026

    This month, Alexandre Sieira and Adrian Sanabria open with a story that punctures some of the AI hype: most of the recent headlines about AI agents "escaping" and hacking other organizations trace back not to superintelligent models, but to a single misconfiguration at Irregular, the third-party firm Google, Meta, Anthropic, and OpenAI used to safely test their models. By Adrian's tally, four of the seven known incidents stem from that one mistake, and only OpenAI's were genuine sandbox escapes. It's a textbook TPCRM lesson: shared responsibility doesn't mean your vendor is getting their half right.Story two is the Manchester Airports Group breach, where attackers abused a read-write API key for Iterable, a marketing tool used just to track email clicks, that sat in the site's client-side JavaScript for four years despite Iterable's documentation explicitly saying never to expose it. Alexandre uses the case to highlight the "hidden third parties" holding your executives' travel data (airports, airlines, hotels) that never pass through your TPCRM process.The episode closes with legal AI company Harvey's shift from frontier models to cheaper open-weight models, which Alexandre uses to untangle two decisions that often get conflated: which model you use, and where its inference runs. Each carries its own jurisdictional, security, and provenance risks, from where your data gets processed to whether a model's training data was poisoned (research suggests as few as 250 documents may be enough). The pair also dig into TypeSafe's new Jev model, built for calibrated decisions rather than text generation. It's hundreds of times faster and cheaper for classification tasks like spotting malicious traffic or routing prompts between models, but its "hallucination-free" claim doesn't hold up: a confidently wrong answer is still wrong, and harder to spot.Links:-Story 1: https://cyberscoop.com/irregular-ai-s...-Story 2: https://scotthelme.co.uk/no-hacking-r...-Story 3: https://www.bloomberg.com/news/articl...-The annual Tenchi conference is coming up soon: https://www.tenchisecurity.com/en/com...-The creator of Jev presents his creation at an AI conference and explains why LLMs aren’t good at automation    • Jev CEO: I made ChatGPT, now I'm building ...  -Alexandre’s LinkedIn post with his prediction about Jev https://lnkd.in/p/dAXiGdye

  3. Sep 15

    Episode 21 | September, 2026

    In this episode of the Alice in Supply Chains Podcast, hosts Adrian Sanabria and Alexandre Sieira discuss the latest news related to third party cyber risk management. - Story one focuses on a wild BGP hijacking incident that compromised updates for a vendor that makes hypervisor management software (Virtualizor by Softaculous). Attackers were able to take over part of hosting company Hetzner’s IP space, create a valid TLS certificate for Softaculous’s website and serve up an identical copy of the website. Softaculous didn’t cryptographically check the integrity of updates, so the attack succeeded in compromising customers. - In story two, the Financial Stability Board, an organization created by the G20 after the 2008 financial crisis, warns of risks from AI. The risks were twofold: the AI bubble and threats directly from rogue AI agents. - In story three, Alexandre shares an observation that, at least in Latin America, financial institutions are writing inside-out scanning into their MSAs. - Finally, in story four, UK politicians propose to amend the Cyber Security and Resilience bill to expand restrictions on tech suppliers seen as ‘risky’. A troubling addition aims to remove transparency at the same time. Links - for reference: Story 1: https://arstechnica.com/security/2026/09/well-executed-bgp-attack-uses-hijacked-ips-to-infect-real-networks/Story 2: https://therecord.media/cyber-risk-from-frontier-ai-most-immediate-concern-to-global-financeStory 3: No linkStory 4: https://therecord.media/uk-technology-national-security

  4. Aug 19

    Episode 20 | August, 2026

    Recording in person for the first time — from a scorching Las Vegas during Black Hat week — Adrian and Alexandre open with Adrian's BSides Las Vegas talk debunking the "60% of breached small businesses fail" myth. Story one covers the "trust me bro, I'll eventually get you a SOC 2" trend: contracts that let vendors start business now and promise a SOC 2 within a year. The guys argue this inverts the whole point — SOC 2 isn't a control, it's an assurance mechanism, so deferring it means hiring first and assessing later. With compliance-automation vendors promising push-button SOC 2s and the economic incentives that produced Delve unchanged, expect more rubber-stamp mills. Story two heads to Brazil, where CMSW Solutions — the first PSTI (PIX payment infrastructure provider) publicly breached in last year's wave of nine-figure heists — passed the Central Bank's tough new requirements, then exited the PSTI business anyway. The lesson for TPCRM teams: "we meet the compliance requirements and no more" is a trap, because sometimes (moving hundreds of millions daily) the regulatory floor sits far below actual risk, and sometimes (a drywall contractor facing CMMC) far above it. Story three was the talk of Black Hat: Hugging Face announced it was hacked by an AI agent, OpenAI raised its hand to claim the oopsie, and Anthropic then admitted its agents had escaped sandboxes and done the same. We discuss the uncomfortable questions: where's the CFAA energy that gets aimed at human researchers, and why do we tolerate two-nines availability and sloppy containment from vendors whose entire pitch is how dangerous their technology is? Practical advice: hold AI labs to a higher vendor standard, and consider running inference through hyperscalers for better infrastructure, IAM, and data residency guarantees. The final story is one for the "wouldn't even occur to a sci-fi writer" file: outside counsel defending Brazilian cosmetics giant Natura embedded a prompt injection in a court filing, attempting to manipulate the judiciary's automated document processing. The judge found the firm litigated in bad faith, applied the maximum 10% misconduct fine, and referred the lawyers to the bar. The TPCRM takeaway: your third parties' AI use isn't just a data-leakage question anymore — review whether your contracts with outside counsel (and everyone else) cover willful acts like this.Links: Story 1: https://www.linkedin.com/posts/kelsey-waters_ive-now-seen-this-twice-enterprise-buyers-share-7483510195877240832-nZFg/Story 2: https://www.linkedin.com/posts/cmsw-tecnologiafinanceira-infraestruturafinanceira-share-7487486750743969794-_xQi/Story 3: https://kaynemcgladrey.com/blog/when-the-arsonist-sells-fire-insurance/Story 4: https://brazileconomy.com.br/empresas/2026/08/natura-e-condenada-pela-justica-por-litigancia-de-ma-fe-e-manipulacao-de-ia/Adrian’s BSides Las Vegas talk: https://youtu.be/K9tDYG4dMSY?t=609

  5. Jul 14

    Episode 19 | July, 2026

    In the July 2026 episode of Alice in Supply Chains, Adrian and Alexandre dig into Verizon's first-ever Breach Impact Study, a companion to the DBIR built on roughly 70,000 cyber insurance claims (accepted and rejected) contributed via Cyber AcuView between 2019 and late 2025. The medians-only approach (no averages allowed!) makes this one of the most honest looks at breach costs yet, and the findings vindicate a recurring theme of the show: availability, not confidentiality, drives losses. Business interruption has the highest median claim (~$90K) and grew 51% year over year, while regulatory fines turn out to be the smallest loss category — so maybe stop leaning on "the regulator will fine us" to justify security budgets. Software supply chain incidents are just 2% of claims but punch far above their weight, with immediate impacts more than double the rest of the dataset and extreme losses topping $100 million — cases that overwhelmingly hit policy caps, meaning the true economic damage is even higher. Adrian connects this to the rise of cascading breaches (Trivy → LiteLLM → everyone downstream), and the pair discuss why this kind of concentrated, systemic risk terrifies insurers and may reshape coverage terms. Story two is Gartner's new "Mastering TPCRM" document series, which Adrian and Alexandre praise as unusually prescriptive and genuinely useful: frameworks for roles and responsibilities across security, legal, procurement, and — critically — business owners, plus a starter TPCRM policy you can adapt. Alexandre argues the make-or-break factor for any TPCRM program is whether business owners formally own the risks they accept, rather than treating the process as a checkbox that must not slow down contracting. Finally, Schrems III is nigh: with the US Supreme Court's ruling undermining the independence of agencies like the FTC — the very foundation of the EU-US Data Privacy Framework — Max Schrems is poised to strike down transatlantic data transfer agreement number three. Alexandre walks through why the EU has painted itself into a corner (idealistic legislation, total dependence on US cloud, AI, and OS providers), what the $307 billion per year in EU spending on US cloud services means for both sides, and why every company with a transatlantic footprint should be talking to their legal counsel now. Links: Tenchi Conference 2026: Details, tickets, and CFP - https://luma.com/vvzsmej9 Story 1: Verizon Breach Impact Study - https://verizon.com/dbir Story 2: Gartner’s Series on Mastering TPCRM - https://www.gartner.com/en/documents/7907309 Story 3: Schrems III is Nigh - https://kaynemcgladrey.com/blog/when-the-load-bearing-wall-comes-down/

  6. Jun 15

    Episode #18 | June, 2026

    This month, Adrian and Alexandre dig into a mix of vulnerability management fundamentals, supply chain security developments, and the accelerating push for European digital sovereignty. The episode opens with a reality check on AI-assisted hacking: despite the hype around tools like Anthropic's Mythos, the vulnerabilities actually causing damage are still the old, unpatched ones sitting on forgotten assets. Cyber insurance claims data from Coalition, At-Bay, and Cowbell is increasingly backing this up with hard numbers — including an significant increase in ransomware risk for organizations running SonicWall devices. The guys also make the case that smart CISOs are using the AI security scare as political cover to finally get boards to fund the basics. From there, the conversation turns to a brewing conflict between Microsoft and the security research community. After researcher Nightmare Eclipse alleged that Microsoft silently patched submitted vulnerabilities without credit or payment, other researchers piled on with similar stories. Microsoft's aggressive legal threats only made things worse — and now Nightmare Eclipse is dropping full disclosures on Patch Tuesday, with more promised. The episode includes practical guidance on what a healthy vulnerability disclosure program looks like, and what happens when organizations get it wrong. The episode also covers npm's new package provenance features as a meaningful but incomplete step against supply chain attacks like the TeamPCP campaign, plus a look at the growing EU digital sovereignty movement — including one Dutch entrepreneur's experience moving his entire stack off US providers, and why subprocessor relationships mean you might not be as decoupled as you think.

  7. May 19

    Bonus Episode with Special guest Alex Pinto | DBIR 2026

    In this special interview episode, Adrian and Alexandre sit down with Alex Pinto, lead author of the Verizon Data Breach Investigations Report, to walk through the 2026 edition before the broader industry has fully digested it. Pinto explains why the 2026 dataset, with 31,850 incidents and 22,624 confirmed breaches contributed by over 100 organizations in 145 countries — is the most statistically rigorous breach corpus in the industry. Tenchi Security is a 2026 contributor, providing the survival-analysis dataset behind the report's new look at third-party MFA and cloud privilege exposures. Alex Sieira walks through what the curves actually mean: half of MFA findings get fixed in seven days, but 45% of cloud privilege management findings are still open a year after discovery. The conversation digs into the headline shifts: vulnerability exploitation has now overtaken credential abuse as the most common initial access vector. Third-party involvement in breaches has climbed from 30% last year to 48% this year, and the median time to fully remediate CISA KEV findings slipped from 32 to 43 days. will probably be the most-talked-about new section of the 2026 report: Verizon analyzed an anonymized dataset from Anthropic. The data includes analysis of nearly 800 threat actors, maps their prompt activity to MITRE ATT&CK techniques, and cross-references it against MITRE's software database. The DBIR folks immediately think to ask the data: “are attackers using LLMs for novel techniques, or for things every EDR already catches?”The trio close out by debating Sieira's hypothesis that the metric to watch isn't total CVE volume — it's the percentage of vulnerabilities with reliable working exploits, which is the variable AI is most likely to move — and Pinto makes the case that vulnerability management is becoming a crisis-management discipline rather than a dashboard-watching one.References: The 2026 Verizon Data Breach Investigations Report (DBIR): https://www.verizon.com/business/resources/reports/dbir/Sieira and Pinto's RSA 2026 talk on how cloud-hyperscaler UX design impacts security outcomes https://path.rsaconference.com/flow/rsac/us26/FullAgenda/page/catalog/session/1755192044047001WRoaThe Vercel Breach: https://cyberscoop.com/vercel-security-breach-third-party-attack-context-ai-lumma-stealer/The British Library breach write-up Adrian cited as a candid post-incident report (their "Learning Lessons" document): https://www.bl.uk/home/british-library-cyber-incident-review-8-march-2024.pdf Tenchi Security has an article out with the biggest insights related to the report, find it here!

  8. May 18

    Episode #17 | May, 2026

    Episode #17 of the Alice in Supply Chains podcast, with co-hosts Alex Sieira (CTO & Co-founder, Tenchi Security) and Adrian Sanabria (Principal Researcher, The Defenders Initiative), is now live! This month they dig into what happens when one breach detonates a chain of others, and why "we have a SOC 2 report" is no defense when the vendor underneath you gets popped. The featured case is Marquis Software Solutions vs SonicWall, where a 2025 breach of SonicWall's MySonicWall cloud backup service gave attackers everything they needed to break into Marquis, drop ransomware, and exfiltrate data on customers of 74 banks and credit unions. Three layers of lawsuits later (consumers suing the banks, banks pressuring Marquis, Marquis now suing SonicWall), Adrian and Alex use the case to make a point about software liability, the absurdity of "as-is" terms in critical infrastructure, and why bare-minimum vendor diligence and self-attestations will surface during discovery as exhibits against you.Story two is the Trivy supply chain compromise, where TeamPCP turned Aqua Security's open-source container scanner into a credential-harvesting beachhead. After an incomplete credential rotation following an earlier incident, the attackers pushed a malicious binary, dropping an infostealer that ran before the legitimate scan and silently swept GitHub tokens, AWS/GCP/Azure credentials, SSH keys, and Kubernetes tokens out of CI/CD runners and developer machines. The blast radius reached Cisco, the European Commission, Checkmarx, Bitwarden's CLI, LiteLLM, Guesty, S&P Global, and seeded the CanisterWorm npm worm. Alex walks through the "how not to get Trivied" playbook: pin GitHub Actions to commit SHAs, kill long-lived CI/CD credentials in favor of OIDC and ephemeral tokens, compartmentalize CI from CD (ideally on different platforms), shrink your dependency graph, and demand evidence of SAST/SCA and IR practice from every third party whose code ends up in your pipeline.Resources: https://destroyedbybreach.comhttps://kaynemcgladrey.com/compliance-paperwork-wont-save-you-from-a-vendor-breach/https://www.acaglobal.com/industry-insights/sonicwall-cloud-backup-breached-firewall-configurations-compromised/https://www.tenchisecurity.com/en/insights-news/secure-practices-trivy-supply-chain-attackhttps://thenewstack.io/teampcp-trivy-supply-chain-attack/

About

In 2022, Tenchi Security created the Alice in Supply Chains newsletter to share and highlight stories on third-party and supply chain risks and attacks, their impacts on services and businesses, and how the industry is moving forward to manage these risks. Following the meteoric success of the newsletter, we've created a podcast! Every month, hosts Adrian Sanabria and Alexandre Sieira will discuss the top six stories from each monthly newsletter, and their thoughts on the future of third party cyber risk management.