2026 NYS Cybersecurity Conference Day 2

Keynote: How AI Changes Everything and Nothing About Education

Dr. Kristen DiCerbo 

Will generative AI change everything or nothing in education? AI has the potential to fulfill long-described potential for personalized learning. However, the education space is highly resistant to change. There are decades of research about human learning that should be brought to bear on the design of AI-powered education tools, but it is not clear whether they will be. Will AI help us address some of the seemingly intractable challenges that education technology has faced in learning and assessment? Or will those challenges remain unsolved? What new challenges does generative AI introduce? Dr. DiCerbo will explore all of this using the lens of lessons learned in the development, implementation and evaluation of Khanmigo, an AI-powered tutor for students and assistant for teachers from Khan Academy.  

Dr. Kristen DiCerbo is the Chief Learning Officer at Khan Academy, where she leads the content, assessment, design, product management and community support teams. Time magazine named her one of the top 100 people influencing the future of AI in 2024. Dr. DiCerbo’s career has focused on embedding insights from education research into digital learning experiences. Prior to her role at Khan Academy, she was Vice-President of Learning Research and Design at Pearson, served as a research scientist supporting the Cisco Networking Academies, and worked as a school psychologist. Kristen has a Ph.D. in Educational Psychology from Arizona State University. 

Zero Trust Beyond the Hype: Patterns of Success and Stall

John Spiegel, HPE

Cybersecurity strategy often looks compelling in diagrams and guiding principles, but real organizations face complexity far beyond the ivory tower. This session translates insights from more than 200 interviews with practitioners, executives and architects who have led Zero Trust–aligned transformations across global enterprises, highly regulated industries and diverse operational environments. Participants will gain practical, experience-based lessons spanning early identity-centric access control initiatives to modern secure service edge and micro-segmentation programs. The session explores why some efforts deliver measurable risk reduction and increased organizational confidence, while others stall due to the misconception that Zero Trust is primarily a technology deployment rather than an operational and leadership transformation. Attendees will learn how to avoid common failure patterns such as over-reliance on tools without aligning identity, governance and process change.

The discussion highlights why culture, sequencing and executive sponsorship are foundational to sustainable progress. It also provides guidance on securing cross-functional buy-in, from CISOs and IT leaders to finance, compliance, and audit stakeholders. Concrete examples of early wins and pragmatic sequencing will demonstrate how to build momentum while aligning cybersecurity initiatives to broader business objectives. By connecting strategy to real-world implementation, this session equips leaders and practitioners with actionable insights to strengthen security posture, improve resilience and drive long-term operational effectiveness through disciplined Zero Trust execution.

Getting Your Org to DMARC Compliance

John Rossiter, Aspire Technology Partners

Deploying DMARC (Domain-based Message Authentication, Reporting & Conformance) involves implementing an email authentication standard that helps protect an organization’s domain from spoofing, phishing and unauthorized use. The deployment process ensures that all outgoing email is properly validated using existing SPF and DKIM records, and that receiving mail servers know how to handle messages that fail those checks.

Building an Effective AI & Privacy Governance Program: From Policy to Practice

Marcy Stevens, NYS Office of Information Technology Services

Valerie Lubanko, NYS Office of Information Technology Services

This session will survey the legal and policy landscape of AI and privacy in the State of New York as a foundation for building and implementing effective governance programs. The presentation will provide regulatory insight, practical implementation and actionable steps for developing a governance process in your organization.

Key Takeaways include:

• Governance structures and framework overview
• Roles and responsibilities
• Key risks (privacy, bias, security)
• Privacy integration

Data Loss Prevention and AI: Protecting Data in an Era of Intelligent Risk

Bill Carter, Fortinet

Artificial intelligence is rapidly transforming how local governments deliver services, but it is also quietly reshaping how sensitive data can be exposed or stolen. In this thought-provoking session, cybersecurity leader Bill Carter explores why traditional compliance alone cannot protect public-sector data in the AI era and introduces practical strategies for building "AI-Safe" data defenses that enable innovation while safeguarding trust. Attendees will gain fresh insight into modern data loss risks, from accidental AI-driven exposure to evolving exfiltration tactics. Learn how forward-thinking governments can protect what matters most while embracing the future.

Tales from the Trenches: Real-Life Incident Response Experiences for your Next TTX

F. Paul Greene, Harter Secrest & Emery LLP

Daniel J. Altieri, Harter Secrest & Emery LLP

Perhaps the most overlooked acronym in cybersecurity is TTX: tabletop exercise. Incident response is a team sport and must be drilled to be effective. A TTX drill, however, is only as good as its content, and too often organizations use canned scenarios for their drills, ignoring the wealth of real-world experience available to make the TTX fit current threat actor tactics, techniques and procedures.

This presentation leverages more than 15 years of front-line incident response experience in the role of a legal “breach coach” to provide cautionary tales of incident response gone wrong, as well as areas for improvement to look for in your incident response plan and team. We will explore changing ransomware tactics, new risks from state-sponsored threat actors, and avoidable mistakes made by organizations before, during and after they learned of the incident. You will leave with practical material you can use to assess the maturity of your information security program, your incident response plan or your organization’s incident readiness more generally.

ASIA Session 5

Paper: Comparative Analysis of VADER and TextBlob on Financial News Headlines

Paper: A Final Return for OpenBSD Anti-Return-Oriented Programming Mitigations

Identity Governance and Administration: 2026 Imperatives for State Government

Andy Egan, CDWG

We’re covering IGA — what it looks like in state government right now, where it’s headed in 2026, why non-human identities are becoming a serious blind spot, and why identity is a board-level problem now. This session will cover:

• What actually is IGA 101, including the five pillars, and why it’s more than just Active Directory.
• Compliance drivers, including CJIS, HIPAA, Pub 1075, and the federal mandates trickling down to states.
• The cost of getting it wrong, including dollars, trust, and operational disruption. This is the heart of it — six trends reshaping IGA for state government in 2026. Where do you actually start? Maturity model and a 36-month roadmap. Who to talk to and what to say — stakeholder messaging and next steps.

Ransomware and the New Economics of Cybercrime

NYS Forum Panel

Emma Yeager, Gartner

Chris Aiwanse, ePlus

John Spiegel, HPE

Ransomware continues to be the most disruptive cyberthreat facing governments, healthcare, education and critical infrastructure organizations. Modern attacks are fueled by sophisticated criminal ecosystems that combine ransomware-as-a-service, double extortion and cryptocurrency-based monetization. This panel examines how ransomware groups operate today, how attacks unfold in real environments, and what organizations can do to improve resilience, detection, response and recovery.

Is your Data PrAIvacy in Jeopardy?

Lily Plotkin, NYS Technology Enterprise Corporation

Rachel Temple, NYS Technology Enterprise Corporation

Michele Warner, NYS Technology Enterprise Corporation

Jeannine Jacobs, NYS Technology Enterprise Corporation

Dom Marando, NYS Office for People with
Developmental Disabilities

Margaret Vijayan, NYS Office for People with
Developmental Disabilities

Artificial intelligence (AI) is reshaping the data privacy landscape by accelerating data collection, analysis and inference at unprecedented scale. Advanced machine learning models can extract sensitive insights from seemingly benign datasets, increasing re-identification risks and challenging traditional de-identification techniques. At the same time, AI is enabling new privacy-enhancing technologies, including automated data classification, anomaly detection, differential privacy and federated learning. This presentation examines the dual role of AI as both a catalyst for privacy risk and a driver of privacy innovation, highlighting regulatory implications, governance strategies and practical controls organizations must adopt to maintain trust and compliance in an AI-driven environment. Yes, the above was AI generated, but this session will feature human professionals and you, the audience, interacting with the topic while playing the fun and familiar game of “Jeopardy.” The Jeopardy categories, trivia and topics will test your knowledge of how AI is impacting privacy, and our panelists will be on hand to educate all on how this rapidly evolving risk can be addressed.

Reasonable Security in the AI Era

Chirag Arora, GlobalCISO Leadership Foundation

Antony Haynes, Albany Law School

Curtis Dukes, Center for Internet Security

As organizations rapidly adopt AI-driven and autonomous (“agentic”) systems, the definition of “reasonable security” is being fundamentally reshaped. What once served as best practice is now becoming a legal expectation. This session explores how the CIS Critical Security Controls® provide a practical, defensible framework for securing modern AI environments. Bridging cybersecurity, legal standards and AI governance, it examines how courts, regulators and industry increasingly evaluate “duty of care” through the lens of established frameworks. Attendees will gain actionable insights into securing AI systems, APIs and non-human identities, while addressing emerging risks such as uncontrolled agent behavior, data flows and third-party integrations. The session also highlights how organizations can align cybersecurity strategy with evolving regulatory expectations, transforming security from a technical function into a business and legal imperative.

The Human Attack Surface: Emotional Intelligence and Cybersecurity

Kiran Bhujle, SVAM International Inc.

Julie Bashant, SVAM International Inc.

Organizations invest heavily in firewalls, endpoint detection, and zero-trust architectures, yet the overwhelming majority of breaches still begin with a human decision. A click on a phishing link, a bypassed policy, a password shared under social pressure – the technology stack is rarely the point of failure. The human is.

This session challenges the prevailing “technology-first” orthodoxy in cybersecurity and presents a “human-first” security model grounded in behavioral science, organizational psychology and real-world incident response experience. Drawing on leadership of managed security operations across government, healthcare and education sectors, as well as insights from teaching organizational change at Columbia University, the presenter will examine why traditional security awareness training consistently fails to change behavior, how emotional intelligence (EQ) assessments can predict an organization’s actual breach susceptibility more accurately than technical vulnerability scans, and what a truly human-centered security architecture looks like when implemented at scale.

The session introduces a novel “Human Attack Surface” scoring model — a lightweight, practical rubric that measures organizational security culture across five dimensions: trust dynamics, communication patterns, authority response, stress tolerance and change readiness. Attendees will participate in a live exercise to assess their own organization using this rubric, producing a baseline score they can take back to their teams as the foundation for a human-centered security improvement program. This presentation does not dismiss technical controls; it argues that without understanding the human layer, technical investments deliver diminishing returns. The organizations that get breached are rarely the ones with the weakest firewalls – they are the ones with the weakest cultures.

ASIA Session 6

Paper: Architectural Trust Assumptions in Implantable Cardioverter-Defibrillator Systems: A Conceptual Analysis

Paper: Cybersecurity Intrusion Detection System with Artificial Intelligence and Deep Learning Using Python

The Agentic SOC: AI, Detection, Automation, and Response at Machine Speed

Jeff Baez, Spunk LLC

Threats are moving faster than traditional SOC models can handle. Security teams are drowning in telemetry, juggling disconnected tools, and spending too much time on manual triage and investigation while attackers exploit speed and complexity to gain advantage.  This session introduces the Agentic SOC as the next evolution of security operations: a model that unifies data, tooling, AI, and automation into a single analyst experience to identify, investigate, and stop threats before they impact the business. Using the HAFNIUM campaign as a real-world case study, the talk will show how modern security teams can turn high-volume data into high-fidelity visibility, streamline detection and response, and accelerate action with AI-driven workflows and specialized agents across the security lifecycle.  Attendees will learn how to use AI and automation to reduce analyst burden, improve prioritization, speed investigation, and stop threats at machine speed. The session will also explore how an agentic operating model helps organizations scale security operations while keeping humans in control of the decisions that matter most.

Automate Your Cybersecurity Defenses as Fast as Your Attackers Develop Threats

Carl Mazzanti, eMazzanti Technologies

Cyberattacks are inevitable — the only question is when. Organizations must automate their defenses as fast as bad actors develop threats, leveraging AI-powered tools to stay ahead.

Nation-state and criminal hackers are increasingly using artificial intelligence to craft highly personalized phishing and fraud attacks against organizations of all sizes. And federal/state regulations (ECPA, NY SHIELD Act, CCPA, etc.) create legal compliance obligations on top of security challenges.

Attendees will learn about defensive actions your organizations should take now, including enhancing your Cybersecurity with AI-backed automated and other defense layers, and training your employees to be aware of phishing and other traps.

The Bottom Line: Organizations that work with a trusted MSP can assess your vulnerabilities and keep your defenses current.

Cybersecurity, AI and Quantum: Protecting Critical Infrastructure in a New Strategic Era

Brianna Bace, USTelecom

Christopher Braccia, USTelecom

Robert Mayer, USTelecom

This session examines the evolving cybersecurity risks facing critical infrastructure. Speakers will outline the current threat landscape and U.S. policy environment. The panel will also explore emerging challenges related to artificial intelligence (AI) and quantum computing, as well as how the industry is preparing for rapidly evolving risks. Finally, we discuss how geopolitical tensions in the Indo-Pacific are fostering a new level of urgency across governments and the private sector.

From Clickops to Code: Automating and Securing System Provisioning, Patching and Compliance

Holden Fenner, Red Hat

Most operations teams manage endpoints such as servers or network devices through console wizards, manual checklists and institutional memory. Systems get provisioned by hand, patches are applied on varying schedules through point-and-click tooling, and compliance checks happen after the fact — if they happen at all. It works until an audit asks who changed what, when and why. This session shows what changes when you move system management into a development platform.

• Provisioning as code: declarative, repeatable system builds that eliminate snowflake servers.
• Signed artifacts and version control: automation content that is versioned, tested, signed and promotable, with built-in rollback and deprecation
• Automated patching pipelines: pipeline-driven patching with testing gates, staged rollouts and automatic rollback on failure
• Security and quality scanning: static analysis and vulnerability scanning integrated into the pipeline, catching issues before production.
• Compliance as code: security baselines encoded as automated checks that run continuously, replacing periodic audits with stale reports.

Each topic includes concrete architecture patterns and real-world production examples. Attendees will leave understanding how a platform approach delivers faster provisioning, reliable patching, continuous compliance and a complete audit trail.

Command Under Fire: Running a Cyber Breach When Systems, Time & Certainty All Collapse

Tim Tipton, Arctiq

Most cyber incident exercises fail because they are safe, polite and reversible. Real breaches are none of those things. They unfold with incomplete visibility, competing authorities, legal pressure, public scrutiny, and pressure forcing leaders to make irreversible decisions before certainty exists. This advanced simulation moves beyond the traditional tabletop and places participants inside a live, constraint-driven breach where decisions shape the outcome in real time.

Attendees operate within a multi-stage cyber incident impacting public-sector operations, critical services and external stakeholders. Information is intentionally fragmented across roles; no single participant ever has the full picture. Technical teams see noisy indicators, executives feel operational and reputational impact first, legal faces disclosure clocks and law enforcement introduces evidence preservation constraints that slow response. Unlike scripted exercises, actions in this session carry consequence. Declaring an incident triggers regulatory timelines. Containment choices affect availability. Delayed decisions increase blast radius. The adversary adapts to defender behavior, escalating when hesitation appears and shifting objectives when defenses harden. Participants will experience how controls degrade under pressure rather than simply fail, how authority and communication gaps magnify damage, and how cognitive fatigue quietly shapes risk outcomes.

The session culminates in a structured after-action analysis focused not on blame, but on identifying early collapse indicators, decision bottlenecks and organizational fault lines that attackers exploit. Attendees leave with practical insight into running incident command under uncertainty, coordinating across disciplines, and building cyber response capabilities designed for reality rather than checklists. This training emphasizes resilience, accountability and speed when perfection is impossible.

ASIA Session 7

Paper: Smart Plugs: Security Vulnerabilities and How to Mitigate Them

Paper: Assessing the Security Vulnerabilities of Baby Monitors

Latest Trends in Cyber Resilience

Jonathon Mayor, Cohesity

Cyber resilience is rapidly redefining how organizations prepare for and respond to modern threats, particularly as ransomware and data-centric attacks continue to escalate in frequency and sophistication. Traditional incident response models — centered on detection, containment and remediation — are evolving into more dynamic, data-centric approaches that prioritize speed, flexibility and assured recovery. This session explores the next phase of incident response, where rapid access to multiple, clean copies of data has become a decisive advantage. Advances in backup architecture, immutability and intelligent indexing are enabling organizations to dramatically reduce recovery times while increasing confidence in data integrity. Rather than treating backups as a last line of defense, leading organizations are operationalizing them as an active component of their cyber resilience strategy.

Adaptive Identity & AI Governance: A Modern Security Playbook

Matt Giblin, Sailpoint

In an era defined by AI and digital acceleration, the old paradigms of identity governance are no longer sufficient. The modern threat landscape demands a move from static rules to a dynamic, "adaptive" security posture. The focus must shift to "real-time, dynamic protection," where security can evolve as quickly as the environment it protects.

We will explore how to build an intelligence-driven framework that not only manages identity of all types (human, machines, and AI agents) but actively secures it with continuous monitoring and automated response.

Join us as we map out a practical path from foundational governance to adaptive AI-driven identity security. 

Enabling the Safe Usage of AI-Augmented Software Development

Brett Sagenich, Black Duck

The rapid adoption and evolution of generative AI presents both unprecedented opportunities and risks for software development teams within state and local agencies. Key concerns include cybersecurity, data tuning and pollution, changes in LLM aptitudes over releases and integration with legacy systems. This session will explore how organizations can responsibly leverage generative AI to advance organizational readiness through innovation while adhering to stringent operational constraints.

Fortifying the Future: A CISO's Guide to Data-Centric Security in the AI Era

Rahul Kohli, KPMG

Venoth Lal, KPMG

This session empowers cybersecurity leaders to navigate the complexities of the modern threat landscape by focusing on a data-centric security model. We will dissect three critical, interconnected domains: proactive data security and privacy, comprehensive data security posture management (DSPM), and the secure adoption of artificial intelligence (AI). Attendees will learn how to move beyond reactive measures and build a resilient security framework that protects their most valuable asset: data. We will provide a product-neutral, strategic roadmap for establishing robust governance, gaining deep visibility into data risks, and securely harnessing the power of AI. This session will equip you with actionable strategies to enhance your organization's security posture and ensure demonstrable compliance in a rapidly evolving digital world.

Enabling Artificial Intelligence for Assistive Technologies: Cybersecurity, Ethics, and Fraud

Michael Ridley, OPWDD

Dom Marando, OPWDD

Tamika Black, OPWDD

AI has transformed the landscape of assistive technologies fundamentally altering the lives of people across all three classes of disabilities i.e., sensory, physical, and cognitive. However, challenges remain in ensuring that some of our most vulnerable people are protected from cybersecurity risks, fraud, and exploitative marketing of products. This panel will discuss how the field of assistive technologies has morphed with recent advances in AI. It also discusses cybersecurity, ethics, and fraud associated with assistive technologies. Finally, it presents thoughts on policy, regulation, and technology needs to address some of these issues.

ASIA Session 8

Paper: Attention Allocation Under Cognitive Load: Effects of Task Complexity and Volume on Security Alert Responses  

Paper: Misinformation and Cybersecurity