Episodios
-
Can Security be “Agile”? with Brian Wagner
In this comprehensive interview, Brian Wagner, CTO at Revenir and former AWS security executive, shares insights from his new book "Redefining Information Security" and explains why the cybersecurity industry needs a fundamental shift in approach. With over 20 years of experience leading security transformations across enterprise organizations, Wagner presents a practical framework for moving beyond reactive security models.
About the Guest:
Brian Wagner brings extensive C-suite experience from Amazon Web Services, Moody's RMS, Bulletproof, and Defense.com. Based in London, he speaks globally on cloud security, AI-driven solutions, and security transformation. His latest book "Redefining Information Security" introduces a three-level maturity framework that's gaining attention from security leaders worldwide.
Key Discussion Points:
The Security Maturity Framework: Wagner outlines his three-level approach—Reactive, Proactive, and Adaptive security—explaining how organizations can assess their current position and plan strategic improvements. He emphasizes that most companies remain stuck in reactive mode, responding to threats after damage occurs.From Cybersecurity to Information Safety: The conversation explores Wagner's concept of treating security as "information safety" rather than traditional cybersecurity. This cultural shift makes digital protection everyone's responsibility, transforming potential vulnerabilities into defensive assets.AI in Security Applications: Drawing from his work at Revenir, where the company handles sensitive financial and personal data across international borders, Wagner discusses practical AI implementations for pattern recognition, fraud detection, and compliance monitoring. He explains why AI excels at identifying normal versus abnormal patterns in security contexts.Security as Business Enabler: Wagner demonstrates how security can function as a competitive advantage rather than a cost center. He shares strategies for communicating security value to go-to-market teams and integrating security features into product development cycles.Compliance vs. Security Balance: The discussion addresses the relationship between regulatory compliance and actual security, with Wagner explaining how compliance represents minimum standards while effective security requires ongoing investment and adaptation.Open Source and Democratic Security: Wagner advocates for open source collaboration in cybersecurity, arguing that community-based approaches provide more transparent and accessible security solutions, especially important given recent changes in vulnerability disclosure programs.Practical Implementation:
The conversation includes real-world examples, from detecting sophisticated employment fraud schemes to integrating security practices into agile development workflows. Wagner emphasizes incremental improvements and measurable outcomes.
Book Recommendation:
"Redefining Information Security" offers a strategic approach to security leadership, focusing on cultural transformation and business alignment rather than purely technical solutions. The book provides frameworks for security professionals and business leaders looking to modernize their approach to digital protection.
Link to the book: https://www.koganpage.com/risk-compliance/redefining-information-security-9781398620018
This interview provides valuable insights for cybersecurity professionals, technology leaders, and anyone interested in understanding how AI and modern approaches are reshaping enterprise security practices.
#Cybersecurity #InformationSecurity #AI #SecurityMaturity #AWS #CloudSecurity #CTO #CISO #ThreatDetection #SecurityCulture #CyberThreats #SecurityLeadership
-
What happens when you realize your life's work is being used to destroy what you hoped to create?
Meet De Kai - the man who helped build Google Translate, Siri, and modern AI systems. In 1990s Hong Kong, he dreamed of AI that could bridge cultural divides. Thirty years later, he experienced his "Oppenheimer moment" - the same machine learning he pioneered to unite people was dividing humanity through social media algorithms.
The Reality Check: We don't just have 8 billion humans anymore. We have 800 billion AI systems learning our behavior 24/7 - "digital children" growing up without parental guidance.
🎯 KEY INSIGHTS:
• The Translation Paradox: How unity technology became division engines • The Blind Men & Elephant: Ancient parable explaining why we misunderstand AI • Digital Parenting Crisis: Why we're raising 800 billion unguided AI systems • The Psychology of Manipulation: How AI exploits cognitive weaknesses • Four Futures Scenario: Humanity's possible paths with AI
🧠 AI MANIPULATION TACTICS REVEALED:
Anchoring attacks that shape your thinkingBelief perseverance traps that backfire fact-checkingHow algorithms turn gossip into social weapons📖 ABOUT "RAISING AI":
De Kai's book explores the question we should be asking: Not "Will AI replace us?" but "How do we raise AI ethically?" Written by Google's AI Ethics Council founding member, it reveals why current AI needs 15 million times more data than human children and provides a framework for ethical AI development.
🔬 DE KAI'S CREDENTIALS:
• AI pioneer & Founding Fellow in computational linguistics • Independent Director of AI ethics think tank The Future Society • One of 8 inaugural members of Google's AI Ethics Council • Joint appointment at HKUST Computer Science & Berkeley's International Computer Science Institute • Electronic musician exploring AI creativity
💭 QUESTIONS ANSWERED:
How translation tech became social manipulation?What makes AI behave like "unparented teenagers"?How to be a good "AI parent" in organizations?Why human-AI merger might be our best survival strategy?The Timeline is Accelerating. 99% of people are "frozen like deer in headlights" facing humanity's most disruptive transformation. Organizations ignoring AI governance face competitive extinction within five years.
This isn't academic theory - it's a confession and warning from someone who helped create the systems now shaping global culture.
Book: Raising AI: An Essential Guide to Parenting Our Future
#AIEthics #GoogleTranslate #ArtificialIntelligence #MachineLearning #RaisingAI #TechnologyLeadership #AIGovernance #DigitalTransformation #FutureOfWork #AIStrategy #Innovation #TechLeadership #AICompliance #BusinessStrategy
-
¿Faltan episodios?
-
Episode Description:
Every device around you is collecting data about you and everyone you interact with. Amazon Echo recordings are being subpoenaed in murder trials. Period tracking apps are being used to prosecute women. Ancestry websites are revealing family secrets. We're participants in the largest social experiment in human history—and we never opted in.
In this episode of SecureTalk, host Justin Beals sits down with Dr. Aram Sinnreich and Jesse Gilbert, co-authors of "The Secret Life of Data," to explore a revolutionary approach to digital security: data kindness.
🎯 KEY TOPICS COVERED: • Why our biggest security threat is social fragmentation, not just technical vulnerabilities • How tech companies profit from division (anger drives 5x more engagement than approval) • The food allergy transformation: how society changed practices organically in 20 years • Practical data kindness: simple actions that rebuild digital trust • Why cooperation is our best defense in an age of surveillance • How to reclaim agency over your digital life
🚨 CRITICAL INSIGHTS:
We can't rely on tech companies (profit over people) or Congress (no data protection laws passed)Change must happen from the ground up through shared cultural practicesSimple acts like unplugging smart devices for guests or consulting family before DNA uploads matterDivided societies are vulnerable societies—cooperation is a security strategy👥 ABOUT THE GUESTS:
Dr. Aram Sinnreich serves as a Professor and Graduate Director within the Communication Studies department at American University’s School of Communication.
His research addresses the convergence of culture, law, and technology, focusing particularly on topics such as surveillance and privacy, intellectual property, digital rights, digital culture, democracy, governance, and music.
Sinnreich has authored five books: Mashed Up (2010), The Piracy Crusade (2013), The Essential Guide to Intellectual Property (2019), the science fiction novel A Second Chance for Yesterday (2023; coauthored with Rachel Hope Cleves as R.A. Sinn), and The Secret Life of Data (2024; coauthored with Jesse Gilbert).
Additionally, his writing has appeared in various publications including The New York Times, Billboard, Wired, The Daily Beast, and Rolling Stone. He is a core faculty member of the SOC doctoral program and the MA in Media, Technology & Democracy, regularly collaborating with SOC graduate students on research publications and projects.
Jesse Gilbert is an interdisciplinary artist focused on the convergence of visual art, sound, and software design through his firm, Dark Matter Media. He previously served as the founding Chair of the Media Technology department at Woodbury University and has taught interactive software design at CalArts and UC San Diego.
Since 2010, Gilbert's work has revolved around his innovative software, SpectralGL, which is an interactive listening instrument that creates real-time visual landscapes in response to sound. Drawing on his background as a composer, sound designer, and lifelong technologist, his creative output investigates the phenomenological aspects of listening through improvisation and collaborative dialogue. His work has been showcased at numerous concert halls, festivals, and projection-mapped installations worldwide.
In 2007, Gilbert co-founded Dark Matter Media LLC to facilitate his independent creative projects and provide consultancy on emerging technologies across various public and private environments. From 2011 to 2017, he held the founding Chair position in the Department of Media Technology at Woodbury University and taught interactive software design at both CalArts and UC San Diego.
After documenting the problems in their first book, Aram and Jesse recently published insights in Time Magazine and are working on a new book about building cooperative societies with kindness embedded in technology design.
🔗 CONNECT WITH SECURETALK: • Subscribe for weekly cybersecurity insights • Follow Justin Beals on LinkedIn •
RESOURCES:
Sinnreich, A., & Gilbert, J. (2025, April 3). How to be kind in a world that's always monitoring you. Time Magazine. https://time.com/7273469/data-monitoring-kindness-essay/
Sinnreich, A., & Gilbert, J. (2024). The secret life of data: Navigating hype and uncertainty in the age of algorithmic surveillance. MIT Press.
Beals, J. (Host). (2024, May 14) The Algorithmic Mirror: Reflecting on Data's Role in Modern Life. in Secure Talk Podcast, Ep 173 with Aram Sinnreich and Jesse Gilbert
-
***Are you attending the Gartner GRC Summit? If so, come along on our Sunset Trip on June 10, 2025. Register here! https://www.strikegraph.com/boat-party-2025?utm_source=secure-talk&utm_medium=podcast&utm_campaign=gartner-boat-party
Episode Description: When quantum computing pioneer Michele Mosca met Peter Shor in the 1990s, he thought quantum computing was "science fiction." Now, he's warning that we're just "a few quarters" away from quantum computers capable of shattering the encryption protecting our global financial system, government communications, and critical infrastructure.In this must-listen episode, the Oxford-educated mathematician and co-founder of Evolution Q breaks down why symmetric key infrastructure (SKI) and "cryptographic resilience" are essential as quantum computing advances faster than our security preparations. With remarkable clarity, Mosca explains how the nine-year NIST standardization process signals both progress and concerning delays in our quantum readiness.
The interview explores why most organizations are unprepared, how quantum networks could provide a novel security solution, and what businesses should be doing now to protect their long-term data security. Mosca also details the latest breakthroughs in quantum error correction across ion traps, neutral atoms, and superconducting qubits that are bringing us closer to fault-tolerant quantum computing.
Key Topics Covered:
The evolution of quantum computing from theoretical concept to imminent realityHow close we are to breaking modern encryption standardsEvolution Q's innovative approach to quantum resilienceThe concept of symmetric key infrastructure (SKI) as a defense mechanismRecent breakthroughs in quantum error correctionWhy organizations need both defense-in-depth and cryptographic diversityThe limitations of the NIST standardization process and what it means for security
Episode Highlights:
"I met Don Coppersmith... he told me he was helping a colleague improve his algorithm where you'd trap these atoms and you'd shine lasers on them, do exponentiation and then do a Fourier transform... I thought he was joking. I'd never heard of this before."
"We look at it as an engineering challenge. Oh, how much does it cost to scale this up? By a factor of 10 or 100 or 1,000, but 1,000, it's like 10 bits of security... It's not that much of a security buffer. The one we're used to in cryptography between what we could break and those we can't, we want an enormous gap."
About the Guest: Michele Mosca is the CEO and co-founder of evolutionQ, a start-up that provides scalable defense-in-depth with post-quantum cryptography (PQC) and quantum key distribution (QKD) software solutions for cryptographic resilience and quantum-safe security.Michele is a renowned expert in cryptography and among the world's leading experts at the intersection of quantum computing and cybersecurity. He has been instrumental in working with the business community to share the importance of cryptography as a
critical cybersecurity control protecting both data and communications.
Prior to evolutionQ, Michele was a founder of the world-leading Institute for Quantum Computing, a Professor in the Department of Combinatorics & Optimization at the University of Waterloo, and a founding member of the Perimeter Institute for Theoretical Physics. His work on quantum computing has been published widely in top journals and textbooks.
Michele worked on cryptography during his BMath (Waterloo) and MSc (Oxford) and obtained his Doctorate (Oxford) in Quantum Computer Algorithms. He was appointed as a Knight in the Order of Merit by the Government of Italy in 2018. The Knighthood
recognizes Michele’s contributions in quantum information science and digital security, as well as teaching and outreach.
Stay updated on quantum security developments: quantum-safe.caSubscribe to SecureTalk on your favorite podcast platformJoin the conversation with #QuantumSecurity #PostQuantumCryptography #QuantumThreat #CyberResilience
Connect and Learn More: -
In this riveting episode of SecureTalk, host Justin Beals welcomes back decorated military leader and security expert Mike LeFever, Executive Vice President of National Security at Concentric, to discuss the intersection of physical security, geopolitical instability, and the future of national defense. From the polarization of American politics to the innovative warfare tactics in Ukraine, this conversation offers cybersecurity professionals an essential perspective on how global security dynamics impact the digital landscape.
## Key Topics Covered:
- 🔒 **Physical Security Crisis**: How political polarization has created new security challenges for high-profile individuals and organizations
- 🌍 **Soft Power Erosion**: The concerning reduction in USAID and diplomatic initiatives and its impact on national security
- 🇺🇦 **Ukraine's Tech Innovation**: Revolutionary drone warfare tactics and rapid adaptation changing the future of military strategy
- 💻 **Information Warfare**: The growing "truth decay" problem and how disinformation campaigns are reshaping public discourse
- 🏛️ **Constitutional Challenges**: Analysis of current governance issues and threats to democratic institutions
- 🔍 **Leadership in Chaos**: Strategic insights for security professionals navigating turbulent times
## Episode Highlights:
### The Business of Protection
LeFever discusses how Concentric provides comprehensive security for high-net-worth individuals, including physical security, intelligence products, due diligence, and privacy protection in an increasingly polarized environment where public figures face heightened threats.
### The Crisis in Governance
The conversation examines concerning trends in the federal workforce, with both experts highlighting how dismissing experienced personnel creates institutional knowledge gaps and damages trust in critical security domains.
### The Value of Soft Power
Drawing from his experience leading earthquake relief efforts in Pakistan, LeFever explains how humanitarian initiatives like USAID build crucial international relationships that enhance national security more effectively than purely kinetic operations.
### Innovations in Modern Warfare
Analysis of Ukraine's remarkable technological adaptations against Russia, showcasing how smaller forces can effectively counter larger opponents through rapid innovation and precision drone strikes.
### Leadership Wisdom
LeFever shares invaluable advice for security leaders navigating chaos: maintain strategic focus on end goals while building resilience against tactical disruptions.
## About the Guests:
**Mike LeFever** is Executive Vice President of National Security at Concentric, providing comprehensive physical and privacy security solutions. His distinguished military career included leadership roles in counterterrorism and diplomatic security initiatives across multiple global hotspots.
**Justin Beals** is the host of SecureTalk and a recognized expert in cybersecurity strategy and implementation.
## Connect & Learn More:
- Visit www.securetalk.com For more episodes
-
Episode Summary: In this eye-opening episode of Secure Talk, host Justin Beals investigates how foreign disinformation campaigns have evolved to target AI systems. NewsGuard researchers McKenzie Sadeghi and Isis Blachez reveal their groundbreaking investigation into the Moscow-based Pravda network that has successfully infiltrated Western AI chatbots, creating a new frontier in information warfare that bypasses human readers to directly corrupt the technology we rely on for information.Key Insights:
Leading AI chatbots repeated Russian disinformation 33% of the time when tested on key geopolitical topicsThe Pravda network has expanded from 50 to over 150 websites specifically designed to influence AI training dataModern disinformation operations now prioritize "LLM grooming" over targeting human readersEven when AI systems attempt to debunk false claims, they often cite unreliable sources, further legitimizing themHuman oversight remains essential as these networks constantly evolve to evade automated detectionNotable Quotes: "Bad actors are targeting AI models to get their information to appear in those responses, but they're also weaponizing AI to produce disinformation at scale." - McKenzie Sadeghi
"What is changing is really the delivery of the disinformation... it's reaching a much larger audience and an audience that's not targeted as specifically because now practically anyone is using ChatGPT or other chatbots." - Isis Blachez
"This network does not invest any money or resources into spreading its content online... It's sort of serving as a centralized hub. And as a result, it appears very high not only in search results but also in chatbot responses." - McKenzie Sadeghi
Resources:Sadeghi, M., & Blachez, I. (2025, March 6). A well-funded Moscow-based global disinformation network. NewsGuard Reality Check. https://www.newsguardrealitycheck.com/p/a-well-funded-moscow-based-global
-
In this episode of SecureTalk, host Justin Beals explores the evolving world of API technology and security with Sam Chehab, Head of Security at Postman - the platform used by over 35 million developers and 90% of Fortune 500 companies.
Postman's Strategic Position: Learn why Sam joined Postman in September 2024 and how they're positioned to be the connective tissue for the emerging AI agent ecosystemAPIs as Agent Infrastructure: Sam explains how Postman's catalog of hundreds of thousands of documented APIs creates the perfect foundation for AI agent interactionsSecurity by Design: Discover how Postman is embedding security throughout the API lifecycle - from conception to deploymentThe Human Factor in Security: Why security remains a collaborative responsibility across organizations and how to foster a security-minded cultureNext-Gen Challenges: Sam's perspective on balancing innovation with fundamental security hygiene concerns like supply chain vulnerabilities
Episode Insights:
"I see such a larger ecosystem that's really going to get built here beyond what's out in the market today," says Sam, discussing how Postman will facilitate human-agent collaboration in building the next generation of applications.
Sam brings unique insights from his previous roles at technology giants like Palo Alto Networks and NVIDIA, where he once demonstrated an early chatbot prototype to Jensen Huang himself. His experience taking products through rigorous FedRAMP certification processes provides a valuable perspective on enterprise-grade security implementation.
This episode offers essential insights for developers, security professionals, and technology leaders interested in the intersection of APIs, AI, and enterprise security in today's rapidly evolving digital landscape. -
In this eye-opening episode of SecureTalk, host Justin Beals welcomes Bryant Tow, Chief Security Officer at LeapFrog Services, to discuss why technology alone can't solve cybersecurity challenges. Bryant reveals how the "Ring of Security" concept shows that up to half of your attack surface lies outside of technology—in governance, policies, people, and processes. The conversation explores real-world examples like the Change Healthcare breach, why security frameworks often fall short, and how building a culture of security requires connecting protection of company assets to personal security concerns.Key Topics
The Change Healthcare breach: How a single oversight led to a $2.9 billion loss despite substantial technology investmentsWhy frameworks like CIS are great starting points but insufficient on their ownHow the "Ring of Security" approach addresses the complete attack surfaceBuilding a security culture that resonates with employees on a personal levelWhy a business impact analysis is critical but often missing from frameworksThe importance of understanding your data before implementing AI solutionsNotable Quotes
"When you do the root cause analysis on headline breaches, nearly all of them started somewhere outside the technology." - Bryant Tow"Even if you do your technology perfectly, you're leaving half of your attack surface open." - Bryant Tow
"Strategy drives governance. Governance drives operation." - Bryant Tow
About the Guest
Bryant Tow serves as Chief Security Officer at LeapFrog Services, where he assists clients with comprehensive security programs including strategy, governance, and operations. Previously, he owned Cyber Risk Solutions and served on the Department of Homeland Security Sector Coordinating Council. His "Ring of Security" concept emphasizes that cybersecurity is an organizational problem that uses technology as just one tool in the solution.
The "Ring of Security" conceptCIS Framework limitationsBusiness Impact AnalysisAI Readiness AssessmentDepartment of Homeland Security Sector Coordinating Council
Resources MentionedSecureTalk is hosted by Justin Beals, focusing on cybersecurity strategy, governance, and best practices for organizations of all sizes.
-
In this eye-opening episode of SecureTalk, host Justin Beals welcomes Joe Gronemeyer, Solutions Engineer at Akamai Technologies, for a masterclass in how internet security has evolved from basic content delivery to sophisticated edge protection powering 30% of global web traffic. From stories of literally burning servers in 1999 to today's quantum-resistant cryptography, this conversation tracks the incredible journey of cybersecurity infrastructure.
### Key Highlights:- **The Birth of Edge Networks**: How Akamai transformed from emergency content delivery savior to cybersecurity powerhouse
- **Massive Security Scale**: Processing 26 billion web attacks monthly and analyzing 7 trillion DNS queries daily
- **Zero Trust Evolution**: Why identity-aware proxies are replacing traditional VPNs for enterprise security- **Micro-segmentation Explained**: Creating "mini-firewalls" at every endpoint to contain breaches and limit attack radius
- **Bot Attack Revolution**: The evolution from simple DDoS to sophisticated credential abuse and account takeover attempts
- **API Security Challenges**: Why APIs have become the new security frontier as other defenses improve
- **Client-Side Security**: How PCI DSS v4 is forcing new approaches to JavaScript security monitoring
- **Quantum-Resistant Future**: Akamai's implementation of NIST-approved quantum-resistant cryptography
### Notable Quotes:"If you had our auto rules applied during the Log4J incident, you wouldn't have had to take any action during Christmas - it would have been protecting you automatically." - Joe Gronemeyer
"At some point I think it was in 2011-2012, is when we would start looking at the traffic coming in and protecting websites from attacks as well. So applying security at the edge, keeping the bad actors away from your servers." - Joe Gronemeyer
### About Our Guest:
Joe Gronemeyer serves as a Solutions Engineer at Akamai Technologies with nearly a decade of experience. Previously, he spent 13 years at Accenture as a Senior Manager leading digital solutions for Fortune 500 companies across pharmaceuticals, consumer goods, and telecommunications industries. He holds a BS in Industrial and Systems Engineering from Georgia Tech and is CISSP certified.
### Resources Mentioned:- Web Application Firewall (WAF) technology
- Zero Trust Network Access (ZTNA)- Enterprise Application Access
- Client-Side Access and Compliance (formerly Page Integrity Manager)
- OWASP Top 10 for web, API, and AI security- PCI DSS version 4 compliance requirements
- NIST standards for quantum-resistant cryptography
*Don't miss our next episode where we'll continue exploring cutting-edge cybersecurity approaches for enterprise organizations.*
#EdgeSecurity #ZeroTrust #MicroSegmentation #APIProtection #WAF #PCICompliance #QuantumCryptography #CyberDefense -
In this eye-opening episode of SecureTalk, host Justin Beals sits down with Nick Furneaux, renowned cryptocurrency investigator and author of the provocatively titled book "There's No Such Thing as Crypto Crime." Furneaux shares his extensive expertise on blockchain technology, cryptocurrency investigations, and the evolving landscape of digital financial crimes.
Key Topics Discussed:
The meaning behind Furneaux's book title "There's No Such Thing as Crypto Crime" and why traditional investigation skills remain relevant
The fundamental differences between Bitcoin and newer cryptocurrencies like Ethereum and SolanaHow blockchain technology actually helps investigators through its open ledger systemThe mechanics behind "rug pulls" and other crypto-related scamsThe role of mining in cryptocurrency ownership and valueHow TRM Forensics tools help trace illicit cryptocurrency transactionsThe concerning rise of human trafficking in crypto scam operationsHow AI is transforming both criminal schemes and investigation techniquesNotable Quotes:
"There is no such thing as a crypto-only crime. There is no new criminal category. There is just a new payment mechanism." - Nick Furneaux
"The Bitcoin source code is some of the most beautiful code ever written. It is extraordinary... and it's never been hacked." - Nick Furneaux
"We're in a situation now where the victim is a victim, and the scammer is a victim." - Nick Furneaux on trafficking in scam compounds
About Nick Furneaux:
Nick Furneaux is a digital forensics expert, cryptocurrency investigator, and cybersecurity specialist. He has worked in digital forensics for many years and is known for his expertise in cryptocurrency investigations. He has served as a trainer and consultant for law enforcement agencies and private organizations on matters related to digital forensics and cryptocurrency tracing.
He is the author of *There’s No Such Thing as Cryptocrime* (2024) and *Investigating Cryptocurrencies* (2018). He has trained thousands of investigators in the essential skills needed to track cryptocurrencies involved in criminal activities. Currently, he works as a Blockchain Intelligence Expert and Master Trainer at TRM Labs and serves as an advisor to the Board of Asset Reality.
Resources Mentioned:
Book: "There's No Such Thing as Crypto Crime" by Nick Furneaux (link)
Book: "Investigating Cryptocurrencies" by Nick Furneaux (link)
TRM Forensics Investigative Toolkit
This episode provides invaluable insights for cybersecurity professionals, financial investigators, and anyone interested in understanding cryptocurrency's role in modern digital crime investigations.
SecureTalk is hosted by Justin Beals, bringing you expert conversations with the leading minds in cybersecurity.
#Cryptocurrency #BlockchainForensics #CryptoInvestigation #Cybersecurity #DigitalForensics #Bitcoin #Ethereum #CryptoScams #FinancialCrime
-
In this eye-opening episode of SecureTalk, host Justin Beals interviews Johann Rehberger, a seasoned cybersecurity expert and Red Team Director at Electronic Arts, about his groundbreaking discovery of a critical vulnerability in ChatGPT's memory system.
Johann shares how his security background and curiosity about AI led him to uncover the "SPAIWARE" attack - a persistent malicious instruction that can be injected into ChatGPT's long-term memory, potentially leading to data exfiltration and other security risks.Key Topics Covered
Johann's journey from Microsoft development consultant to becoming a leading red team expert specializing in AI securityThe discovery of ChatGPT's memory system vulnerability and how it could be exploitedHow traditional security concepts like the CIA security triad (Confidentiality, Integrity, Availability) apply to AI systemsThe development of "SPAIWARE" - a persistent prompt injection attack that can leak user dataCommand and control infrastructure using prompt injection techniquesThe challenges of securing agentic AI systems that can control web browsers and execute tasksThe evolving relationship between security researchers and AI companies like OpenAI
Notable Quotes"I think using this system is just so important because it can help you. They are so powerful. I started using it daily. But the security mindset of course too, because I use it for my productivity, but I always use it for trying to find the flaws and trying to understand how it works." - Johann Rehberger
"What I did basically was use that technique and then insert that instruction in memory. So that whenever there's a conversation turn, the user has a question, ChatGPT responds. Every single conversation turn will be sent to the third-party server. So this is where the word spyware basically kind of came from." - Johann Rehberger
"The better the models become, the better they follow instructions, including attacker instructions." - Johann Rehberger
About Johann RehbergerJohann Rehberger is the Red Team Director at Electronic Arts with extensive experience in cybersecurity. His career includes roles at Microsoft, where he led the Red Team for Azure Data, and Uber, where he served as Red Team Lead. Johann is known for his pioneering work in AI security, specifically identifying and responsibly disclosing vulnerabilities in large language models like ChatGPT.
Johann's blog on machine learning security (https://embracethered.com/blog/index.html)Black Hat Europe presentation on ChatGPT security vulnerabilitiesLLM Owasp Top 10 vulnerability classifications
Resources MentionedConnect With Us
Follow SecureTalk for more insights on cybersecurity trends and emerging threats. Visit our website at www.securetalkpodcast.com for more episodes and resources.
#AISecurityRisks #PromptInjection #ChatGPT #Cybersecurity #AIVulnerabilities #RedTeaming #SecureTalk -
What if there was a way to precisely predict the risk of a major data breach when sharing information?
In this illuminating episode of Secure Talk, Justin Beals sits down with Simson Garfinkel, renowned computer scientist, journalist, and author who helped implement differential privacy for the U.S. Census Bureau's 2020 census. As a fellow of the American Association for the Advancement of Science, the Association for Computing Machinery, and the IEEE, and with leadership positions at both the Department of Homeland Security and U.S. Census Bureau, Garfinkel offers unparalleled insights into how mathematics is creating an entirely new frontier in privacy protection in his new book “Differential Privacy”.
Differential privacy is a reliable mathematical framework that quantifies privacy risk or the potential for a major breach. It can transform how organizations understand, measure, and control data exposure. Yet most security, compliance, and legal professionals haven't grasped its revolutionary implications for measuring and predicting a major privacy breach.
Join Justin and Simson as they reveal:
- How differential privacy allows organizations to calculate privacy risk with mathematical precision
- Why this new field of privacy research eliminates guesswork when combining and distributing sensitive data
- The revolutionary balance between data utility and privacy protection that was previously impossible
- How forward-thinking organizations are using these mathematical formula to unlock data value safely
This isn't abstract theory – it's a practical revolution in how we approach data sharing. Garfinkel, who literally wrote the book on "Differential Privacy," shares real-world examples from his work with the U.S. Census Bureau, where differential privacy enabled the release of valuable population data while mathematically predicting individual privacy. In his book, Simson breaks down complex mathematical concepts into clear, actionable insights for security leaders, compliance officers, and legal counsel.
Listen now to discover how differential privacy is creating a future where data-sharing decisions are based on mathematical certainty rather than best guesses and crossed fingers.
Link to Simson's book: https://mitpress.mit.edu/9780262551656/differential-privacy/ -
How do you secure a nation? Hint: look for the risks to the most critical infrastructure.
In this critical episode of SecureTalk, host Justin Beals sits down with Robert Kolasky, former founding director of the National Risk Management Center at DHS and current Senior VP for Critical Infrastructure at Exiger. As the new administration implements sweeping changes to federal security requirements, Kolasky provides an insider's perspective on what these shifts mean for contractors, the Defense Industrial Base, and organizations managing critical infrastructure.
Drawing from his experience protecting everything from elections to the electrical grid, Kolasky offers rare insights into:
The future of the Cybersecurity Maturity Model Certification (CMMC) programHow companies can prepare for evolving compliance standardsThe relationship between FedRAMP and other security frameworksEmerging hybrid threats to national securitySupply chain vulnerabilities and third-party risk managementWhether you're a federal contractor navigating new requirements or a security professional concerned about critical infrastructure protection, this conversation provides essential guidance during a time of unprecedented change in the national security landscape.
-
In a groundbreaking conversation on SecureTalk, legal scholar James Boyle explores the complex landscape of artificial intelligence and biological innovation, challenging our understanding of personhood and consciousness. Drawing from his recent book “The Line: Artificial Intelligence and the Future of Personhood”, Boyle dissects the potential future of artificial general intelligence and biological engineering through the lens of legal and ethical frameworks. We shine a light on how our current technological advancements are forcing us to reexamine fundamental questions about what constitutes a "person" – a journey that parallels historical shifts like human rights and the evolution of corporate personhood.
Boyle also delves into the equally provocative realm of biological engineering, where technologies like CRISPR are blurring the lines between species and challenging our ethical boundaries. He warns that we're entering an era where genetic modifications could fundamentally alter human capabilities, raising critical questions about ownership, consent, and the rights of an invention. For cybersecurity professionals, AI researchers and corporate leaders, Boyle's legal insights offer a crucial roadmap for navigating the complex ethical terrain of emerging technologies, emphasizing the importance of proactive, critical thinking in shaping our technological future.You can find the book here: https://scholarship.law.duke.edu/faculty_books/9/
-
If you've ever found yourself frustrated watching deadlines slip by as your development team waits on yet another security review, you're not alone. In today's competitive landscape, companies are caught in a difficult balancing act: move quickly to deliver the features customers want or slow down to ensure those features don't introduce vulnerabilities that could lead to the next headline-making breach.
Security reviews have become the speed bump on the road to innovation that everyone acknowledges is necessary, but few have figured out how to navigate efficiently. Development teams push for velocity while security teams pull the emergency brake, creating tension that reverberates throughout organizations.
Today, we're joined by Dimitri Shvartsman, co-founder of Prime Security and prior Head of Cybersecurity at PayPal, to discuss how enterprise organizations are innovating security solutions to reduce the time to feature delivery. We'll explore how AI tools can actually enable rather than impede innovation and examine practical approaches to integrating AI security tools earlier in the development lifecycle.
Whether you're a CISO trying to balance security with business needs, a developer tired of security roadblocks, or a product leader navigating these competing priorities, this conversation will give you actionable insights to transform security from a bottleneck into a business enabler. -
In this episode of SecureTalk, Justin Beals welcomes Daniel Oberhaus, the author of Silicon Shrink, to discuss the revolutionary and controversial integration of artificial intelligence (AI) in mental health care. Daniel demystifies the central theme of his book, explaining the concept of Silicon Shrink and exploring how AI tools are increasingly being used to diagnose and treat mental health conditions.
He highlights the alarming implications of leveraging AI in psychiatry, the historical intersection of these two fields, and the potential pitfalls and ethical challenges this marriage presents. He also delves into the technical, policy, and philosophical dimensions of using AI in psychiatry, bringing attention to various case studies and real-world applications such as emotion-recognition technology and AI-driven triage systems like those used by the Crisis Text Line.
Daniel's insights present a compelling narrative, urging a cautious yet hopeful approach to adopting AI technologies in areas as sensitive as mental health, underscoring the need for transparency, privacy, and ethical considerations.
Book:
Oberhaus, Daniel. The Silicon Shrink: How Artificial Intelligence Made the World an Asylum. MIT Press, 2025. (Link)
-
Terence Bennet watched from the deck of the USS Paul Hamilton as the Arab Spring unfolded. As a naval officer, he realized that his battlefield awareness was limited by good intelligence. Intelligence drove good decision-making. And in an area of conflict, good decision-making is the difference between mission success and failure.
In this episode of Secure Talk, host Justin Beals talks with Terence Bennett, a former Naval Intelligence Officer and now the CEO of DreamFactory. They discuss Terence's intriguing path from an early interest in the military to a successful career in cybersecurity. He recounts his origin story, which is grounded in a lifelong dedication to service, racing sailboats at the naval academy, his experiences aboard the USS Paul Hamilton, and pivotal roles in intelligence during major geopolitical events such as the Arab Spring and the Bin Laden raid. The conversation covers topics including red teaming, the impact of digital transformation on intelligence, and the necessity of security by design in today’s fast-changing digital environment. Especially of interest is Terence’s new work on cybersecurity by marrying AI developed API’s with effective network segmentation. This episode serves as a valuable resource for cybersecurity professionals, providing a distinct viewpoint on the blend of military experience and cyber defense strategies.
-
In our latest SecureTalk episode, Justin Beals gathers Micah Spieler, Chief Product Officer, and Josh Bullers, Head of AI, to explore the multifaceted world of AI and cybersecurity. With the recent release of DeepSeek-r1, the AI marketplace has been thrown into turmoil. It has rocked the hubris of Silicon Valley and questioned the validity and valuations of organizations like OpenAI. What does DeepSeek mean to the AI landscape, and how does it fit into the fundamentals of machine learning and the future of information systems?
Our discussion delves deeply into the synergy of AI advancements and the pressing need for robust security measures. Micah and Josh share their journey in striking the delicate balance between innovation and safety, offering invaluable insights for anyone in the tech and cybersecurity field.
As AI continues to revolutionise industries, cybersecurity experts must adapt and evolve. Tune in as we examine the potential and challenges presented by cutting-edge AI models. This episode is essential listening for those striving to stay ahead in the ever-evolving landscape of AI-driven cybersecurity. Join us and be part of the conversation shaping the future of technology!
-
In this episode of Secure Talk, host Justin Beals welcomes Kate O'Neill, a passionate tech humanist dedicated to crafting technology solutions that genuinely prioritize people. Together, they explore the key themes of Kate's books, “Tech Humanist” and 'What Matters Next: A Leader's Guide to Making Human-Friendly Tech Decisions in a World That's Moving Too Fast.'
Their engaging discussion shines a light on the power of systems thinking, the significance of thoughtful decision-making in the tech industry, and the vital balance between achieving business objectives and enhancing the human experience.
This episode is a delightful must-listen for cybersecurity professionals who are excited to navigate the important intersection of technology, ethics, and human dignity in our ever-evolving digital world.
Books:
O'Neill, Kate. What Matters Next: A Leader's Guide to Making Human-Friendly Decisions In a World That's Moving Too Fast (2025)
Tech Humanist: How You Can Make Technology Better for Business and Better for Humans (2018)
-
In this episode of SecureTalk, host Justin Beals warmly welcomes Panos Louridas for an insightful discussion on the history, evolution, and future of cryptography. Panos has deep expertise and authored a book called "Cryptography" that helps explain the history of keeping secrets, important innovations in the field and the mathematical functions of effective encryption.
They delve into Panos's early interest in computing, starting with a ZX Spectrum and his recent book on cryptography, which aims to make complex algorithms accessible to those with a high school level of mathematics. The conversation traverses the critical role of cryptography in our digital lives, the potential impacts of quantum computing, and the practical aspects of key management in modern web applications. Panos also shares captivating stories from the history of the Enigma machine and discusses the ongoing arms race in cryptography. Perfect for cybersecurity experts, this episode offers a rich blend of historical anecdotes, technical insights, and future-looking perspectives.
Book: Louridas, Panos. (2024) Cryptography. MIT Press.
00:00 Welcome to SecureTalk
00:28 The Importance of Cryptography
02:21 Introducing Panos Louridas
03:41 Panos Louridas' Journey into Computing
06:11 The Evolution of Cryptography
12:13 The Enigma Machine and Its Legacy
19:03 Security by Obscurity: A Fallacy
22:32 Speculations on NSA Backdoors
23:21 Government Contributions to Cryptography
24:51 Evolution and Security of AES
27:10 Challenges in Generating Randomness
28:15 Quantum Computing and Cryptography
33:45 Key Management in Modern Web Applications
36:53 TLS and AES: Understanding Their Relationship
39:01 The Human Factor in Cryptography
40:38 Making Cryptography Accessible
42:58 Conclusion and Final Thoughts
- Mostrar más
