34:55This recording is not available yet.
Data Security – Why Now?
Download resourcesCategories
AI & Emerging TechnologyPrivacy & Data ProtectionData GovernanceData Loss PreventionPrivacy LawGovernance, Risk & LeadershipSecurity Program & Strategy
Sponsored by
About this session
Michael 'Fergo' Ferguson, Principal Strategist for Security Services at Cyera and author of Cyera's DSPM certification, argues that data security has become the industry's most urgent, least mature discipline because control has shifted from on-premise perimeters to distributed SaaS and cloud data nobody fully inventories. He walks the classic data lifecycle (discover, classify, govern, protect, monitor, destroy) and argues each step only works if the one before it did, then contrasts human review (about 60 bytes per second of reading throughput) with an AI classification engine that scans structured and unstructured stores, infers subject-matter context (a name plus a UK address becomes a GDPR-relevant customer record) at roughly 95 to 97 percent precision versus the roughly 58 percent Fergo cites for Microsoft Purview's auto-labeling. He extends the same discovery-first logic to AI risk (shadow AI, agentic AI, third-party LLM integrations) and to DLP, where feeding classification context back into alerts can cut the roughly 40 percent false-positive rate. A long Q&A covers positioning against Microsoft, Netskope and Zscaler, tenant data isolation, onboarding time (roughly one to two weeks to reach high accuracy), on-premise/cloud deployment flexibility, and scan speed at scale (terabytes in days, petabytes in weeks).
In an era of exponential data growth, fueled by GenAI and data democratization, the need for strong data security has never been greater. Join this session to explore how organizations can safeguard sensitive data while still driving innovation and agility, including: Strategies for balancing robust data protection with business enablement in a fast-changing digital environment. Approaches to tackling the unique security challenges introduced by GenAI and expanded enterprise data access. Tactics for aligning key stakeholders and building consensus around data security priorities.
Key takeaways
- Sequence your data security program: discovery must precede classification, which must precede governance and access controls, which must precede monitoring and response; skipping a step breaks the ones after it.
- Don't rely on regex/keyword-based classification alone for policy enforcement; it cannot tell a customer's credit card from an employee's or infer residency-driven compliance obligations the way context-aware AI scanning can.
- Inventory AI exposure the same way you inventory data: shadow AI apps, agentic AI with index access, and SaaS features quietly calling third-party foundational models all need to be mapped before you can govern them.
- Feed classification context back into DLP alerts to cut false positives (roughly 40 percent in unadjusted systems) by distinguishing a legitimate action on sensitive data from an actual policy violation.
- When evaluating a data security vendor, ask about tenant isolation, on-prem versus cloud deployment options, and time-to-accuracy (expect roughly one to two weeks of ingestion before classification stabilizes).
Speakers
Michael Fergo
Michael Ferguson, known as Fergo, is the Data Security Services Strategist at Cyera. With 20 years of experience in Data, Web, Cloud, and Email Security technologies, he specializes in Data Security and Zero Trust initiatives. As the author of… Read moreRead less
Michael Ferguson, known as Fergo, is the Data Security Services Strategist at Cyera. With 20 years of experience in Data, Web, Cloud, and Email Security technologies, he specializes in Data Security and Zero Trust initiatives. As the author of Netskope’s SASE Accreditation, Cyera DSPM Security Architect accreditation, Fergo has played a key role in shaping data security strategies for multinational organizations across the UK, Germany, Australia, and the USA.
