The Virtual CISO

• The AI Supply Chain: Risks Beyond Your Organization

  Your AI system is only as secure as the ecosystem it depends on.From third-party APIs to pretrained models, most organizations are now relying on external AI components they don’t fully control and that’s where hidden risks live.In this episode of  Securing AI, we explore how AI supply chain risk is reshaping how CISOs and business leaders think about assurance, dependency management, and vendor trust.How do you validate the integrity of a model you didn’t build?What happens when your “trusted” AI partner quietly integrates another model trained on sensitive data?And what does supply chain transparency look like in a world where even algorithms have dependencies?Join me as we unpack the expanding risk perimeter of AI ecosystems and why trust in the age of machine intelligence must extend far beyond your own organization.#AI #CyberSecurity #RiskManagement #SupplyChainSecurity #SecurelySpeaking #LeadershipWe would love to hear from you. Reach out: [email protected]

  --------  

  5:12

  --------

  5:12
• Governance in the Age of AI: Who Owns the Risk?

  CISOs, boards, and product teams are all racing to embrace AI, but when something goes wrong, who takes responsibility?In this episode of Securely Speaking: Securing AI, we explore the evolving challenge of AI governance, where accountability, ethics, and innovation collide. As organizations deploy AI faster than they can regulate it, traditional governance models are struggling to keep pace.From boardroom oversight to engineering decisions, this episode unpacks:Why AI governance isn’t just a compliance issue , it’s a business risk issueHow ownership of AI risk is shifting across the enterpriseThe emerging role of CISOs and cross-functional governance councilsWhy accountability must be designed in, not bolted onAI governance isn’t about slowing innovation, it’s about making sure the systems we build can be trusted.#CyberSecurity #AI #Governance #RiskManagement #Leadership #SecurelySpeaking

  --------  

  4:48

  --------

  4:48
• AI and Data Privacy: When Sensitive Data Trains the Algorithm

  AI systems are only as ethical and secure as the data that trains them. But what happens when that data includes sensitive or regulated information?In this episode, we unpack the growing tension between innovation and privacy, exploring how organizations can responsibly harness AI without crossing legal or ethical lines.We’ll cover:The hidden risks of training data exposureWhy anonymization isn’t always enoughHow privacy laws like GDPR and CCPA are evolving for AIWhat leaders can do to build transparency and trust into their modelsThe future of AI trust depends on how we handle data today. Let’s talk about what responsible innovation really looks like.

  --------  

  6:39

  --------

  6:39
• Securing AI Models: Protecting the Brain of the Machine

  In SaaS, data was the crown jewel. In AI, the model is the brain. If you can’t secure it, you can’t secure your product.In this episode of Securing AI, we move beyond data security and step directly into the core of AI risk: the model itself. While many teams focus on infrastructure and compliance, most breaches in AI won’t come from the cloud platform, they’ll come from poisoned data, manipulated prompts, stolen model weights, and unseen model behaviour.Listen and learn about: - Model theft, exfiltration, and IP risk : when your competitive edge becomes someone else’s asset- Training data poisoning & prompt manipulation : how adversaries reshape outputs without touching your systems- Shadow experimentation: internal experimentation without governance or guardrails- Why “securing AI” is not the same as securing an applicationThis episode challenges you to treat model security as a direct business risk because if the model can be influenced, every decision it makes can be compromised.#ai #SecuringAI #llm #gemini #chatgpt #compliance #anthropicai

  --------  

  8:11

  --------

  8:11
• SOC 2 for AI : Can Traditional Frameworks Keep Up?

  SOC 2 wasn’t written for AI. But customers still demand proof of trust. The question is: are we adapting our frameworks, or just checking boxes that no longer fit? In this episode of Securing AI, we unpack the tension between legacy compliance frameworks and modern AI-driven products. SOC 2 remains the gold standard for SaaS trust, but when models learn, evolve, and operate autonomously, traditional control criteria start to fall short.We explore:- Why SOC 2’s Trust Services Criteria must be reinterpreted for AI systems- How to map AI risks like model drift, data lineage, and API dependency to existing controls- The danger of claiming compliance without addressing model transparency, privacy, and third-party AI providers- What founders, CISOs, and security leaders must do to maintain credibility with enterprise buyersThis isn’t about passing an audit, it’s about proving trust in an era where AI decisions are no longer fully explainable.Listen in if you're building, deploying, or governing AI products and want to turn compliance from a checkbox into a strategic trust advantage.#ai #compliance #podcast #foryou #security

  --------  

  4:37

  --------

  4:37

Altri podcast di Tecnologia

Podcast di tendenza in Tecnologia

Su The Virtual CISO