KBKAST

KB sits down with Michael Pastor, Dean for Technology Law Programs at New York Law School and former General Counsel of New York City Cyber Command, to talk about who is accountable when AI gets it wrong.

They start with Mythos-class tools and what they mean for the legal standard of reasonable security. Michael argues the test is effectiveness and what your people can validate, not cost, and that "I trusted the technology" won't survive discovery.

From there they get into whether AI could help settle disputes between counterparties, how courts might handle highly technical cyber cases, and why Michael wants to retire the phrase "human in the loop" in favour of real oversight by people with context.

They finish on contracts - what general counsel should demand from AI vendors now, why renewals are about to get harder, and why Michael is still optimistic about where it all lands.

About Michael:
Michael Pastor is a Professor of Law from Practice. He serves as New York Law School's Dean for Technology Law Programs, the Director of the James Tricarico Jr. Institute for the Business of Law and In-House Counsel, and Co-Director of the Innovation Center for Law and Technology. He teaches the In-House Counsel Experience Seminar and Drafting Contracts: Technology Transactions.

Professor Pastor previously served as the General Counsel of the New York City Department of Information Technology and Telecommunications (DoITT) and the New York City Cyber Command (NYC3).

Keywords: reasonable security, standard of care, AI liability, general counsel, legal counsel, technology contracts, AI vendor contracts, vendor management, audit rights, breach notification, human in the loop, human oversight, NYDFS, cybersecurity regulation, vulnerability management, Mythos, AI validation, contract renewal, privacy by design, security by design, New York City Cyber Command, CISO, board, KBKast

In the first weeks of March, Omar Santos got early access to the model that became Anthropic’s Mythos. About an hour of testing was enough to convince him cybersecurity had changed for good.

Omar, a Distinguished Engineer in AI Security Engineering at Cisco and co-chair of the Coalition for Secure AI, says his bigger worry is how fast everyone else is catching up. Open weight models are closing in on the frontier labs, so attackers no longer need access to Mythos to find and exploit vulnerabilities.

KB and Omar get into why Cisco won’t send a finding to engineering without proof of an active exploit, how a public open source fix gives attackers a ready-made recipe, and why a library a student abandoned years ago could be running inside your patient records system. Omar wants defenders to go further than finding and fixing faster. That means refactoring code, adding compensating controls, and turning the risk register from an annual exercise into something that runs continuously.

About Omar:

Omar has written more than 25 books, 21 video courses and over 50 research papers, and holds multiple patents. Before Cisco, he served in the United States Marines.

Keywords: Mythos, Anthropic, Project Glasswing, frontier AI, open weight models, AI security, AI in cybersecurity, vulnerability management, zero days, exploit development, open source security, software supply chain, SBOM, OpenEoX, CSAF, CodeGuard, Coalition for Secure AI, CoSAI, Cisco, compensating controls, eBPF, secure by design, zero trust, incident response, threat hunting, continuous assessment, risk register, GRC, CISO, board, KBKast

Most boards approved AI for the productivity gains, and far fewer asked what it would expose.

Dean Saunders says conversations with large organisations usually start with models and end up on data: where it lives, who can reach it, and how quickly an agent can move it once staff have gone home.

KB and Dean talk about shadow AI, approved tools used on personal accounts, and trusted apps that quietly add AI features without security knowing. They also get into why compliance keeps getting mistaken for security, and why regulators now want proof of what you did rather than the policies you wrote.

About Dean:

With more than two decades of leadership experience in enterprise software, cybersecurity, SaaS, channel strategy and go‑to‑market execution across Australia and the broader Asia Pacific, Dean has built a strong track record driving regional growth. His background includes leading high‑performing sales teams, developing partner ecosystems and supporting technology organisations competing in complex markets.

Before joining Forcepoint, Dean held senior commercial and leadership positions at Cyara, Forticode, Unitrends/Kaseya, Teradici, CommVault and Total Defense. His diverse experience spans direct sales, channel and alliances, partner‑led growth and strategic market development, complemented by advisory work focused on improving sales execution and market strategy. He holds an MBA from the University of the Sunshine Coast and has completed extensive professional development in enterprise selling, leadership, key account management and media training.

Keywords: AI data security, data security posture management, agentic AI, shadow AI, data loss prevention, DLP, data discovery and classification, AI governance, compliance vs security, insider risk, Forcepoint, cyber risk, board cyber governance, KBKast

Dr. Ahana Datta Fasel has spent years mapping 30 years of China’s cyber operations, and she opens by dismantling the story most of us have absorbed. China isn’t only hunting our secrets. Sometimes it wants to know what we know about them, sometimes it’s watching its own diaspora, and often the intent behind an operation is genuinely hard to read.

She walks KB through why cyber has become inseparable from live conflict, from Ukraine’s power grids to the CCTVs of Tehran, and why the line between spying and preparing for war keeps moving. She explains the operation sitting inside America’s major telecom networks, where hackers have held access for three years and done nothing with it, and what that patience might mean. We also get into Five Eyes politics, the Huawei decisions, and why the West still doesn’t have a coherent China strategy.

Ahana’s book, Full Stack Spies: Cyber Espionage in the Age of US-China Competition, is out in North America on 1 October. Grab it on Amazon (https://a.co/d/01aovOFW) or from Oxford University Press (https://global.oup.com/academic/product/full-stack-spies-9780197871386), where the code ADISTA5 takes 30 percent off.

About Ahana:

She serves as non-executive director at a leading global digital rights charity, adviser to deep tech start-ups, and a strategic adviser on national security and resilience to the UK body for research, innovation and international partnerships. She has held fellowships at Cornell, Cambridge, Imperial and Durham, and her op-eds have appeared in the FT, Foreign Policy, Euractiv and the Columbia Journalism Review.

Keywords: China cyber operations, cyber espionage, US-China competition, Full Stack Spies, geopolitics, Five Eyes, telecom breach, act of war, Taiwan, national security, hybrid warfare, Huawei

Shane Buckley, CEO of Gigamon, has watched the AI spending story turn in months. Companies that treated token consumption as a badge of honour are now blowing through their AI budget by day three of the month, with the bill hitting earnings reports and boardrooms.

He explains why more tools never meant more security, why he expects rising data centre costs to push token prices up, and why augmentation beats replacement. Then comes the security squeeze. Using his pizza analogy, AI eats the budget just as attackers get their own AI.

About Shane: Shane Buckley is President and Chief Executive Officer of Gigamon, having previously held the title of President and Chief Operating Officer before his appointment as CEO in 2022. He brings more than 25 years of executive management experience to the team with prior roles including CEO at Xirrus, General Manager and Senior Vice President at NETGEAR, President and CEO of Rohati Systems, and COO at Nevis Networks. Shane is a graduate of engineering from the Munster Technological University in Ireland.

Keywords: tokenomics, AI budget, tool bloat, next-generation SIEM, data lake, AI security, augmentation, hybrid AI, on-premise LLMs, cloud adoption, network visibility, vibe coding

Reece takes us behind the scenes of the Instructure incident, where his team fielded 65 instructions in 48 hours, and explains how a catastrophic cyber event actually gets run across insurers, government and thousands of schools at once. KB & Reece get into the hidden cost almost no one budgets for, business interruption, and what the JLR and Change Healthcare cases reveal about how dependent our economy has become on a handful of suppliers.

He also makes the case that AI is quietly splitting security budgets in two, and why that worries him. For any leader thinking seriously about resilience, this one is worth your time.

About Reece:

Reece Corbett-Wilkins is the Head of First Response Australia, and is the Chief Strategy Officer at Atmos. For over 10 years, Reece has focused exclusively on cyber and privacy risk, having helped 1,000+ organisations through some of their most challenging times. Reece is particularly experienced in managing large scale multi-party data breaches, both locally and globally, and ransomware response. Reece works closely with the incident response industry, government, and law enforcement to achieve optimal outcomes for clients.

Keywords: cyber resilience, incident response, business interruption, cyber insurance, critical infrastructure, SPHERE, ransomware, supply chain risk, SOCI reform, AI security budget, CAT event, small business cyber

When Mythos hit in April, the security world split between panic and hype. Sumedh Thakar, President and CEO of Qualys, joins KB to make the calmer case that Mythos accelerated an old threat rather than inventing one. They get into why zero-day vulnerabilities now need zero-day remediation, the misread that Mythos runs on your own code while your vendors use it too, the board conversation it reopened, tokenomics and budget pressure, and how much a CISO should really hand to the machines.

About Sumedh:

Sumedh became president and CEO in 2021. In 2019, he was named president, and prior to that, he was chief product officer, driving the company’s vision of making enterprise security more efficient and disrupting the VM space with integrated capabilities like patch management and cybersecurity asset management. A “product fanatic and engineer at heart,” Sumedh was instrumental in dramatically expanding the original Qualys platform’s scope, integrations, and automations. He also scaled the company’s engineering talent internationally with a global 24×7 follow-the-sun product team. He is a co-inventor of five U.S. patents for cybersecurity technology in Qualys offerings.

Previously, Sumedh was an engineer at Intacct, an early cloud-based financial and accounting software provider. He also worked at Northwest Airlines developing complex algorithms for its yield and revenue management reservation system. He has a bachelor’s degree in computer engineering with distinction from Savitribai Phule Pune University.

Keywords: Mythos, AI security, cybersecurity, zero-day, autonomous remediation, vulnerability management, CISO strategy, board reporting, Qualys, Sumedh Thakar, patch management, true risk, exposure management, AI attacks, first-party code

The tech running hospitals, power grids and government services often still works. Sarah Sloan, Cisco’s Head of Cybersecurity Policy for APAC, joins KB to explain why that’s the problem, not the reassurance it sounds like.

Drawing on a new ASPI report Cisco funded, “Past its use-by-date,” they get into why so much end-of-life tech is still in place (usually budgets and skills, not negligence), how AI and quantum turned a slow-burn risk into an urgent one, and why most organisations still can’t see the ageing gear in their own environment. Plus the upside: why moving early beats being forced by an incident.

About Sarah:

Keywords: end-of-life technology, legacy systems, critical infrastructure, cybersecurity governance, ASPI, Cisco, Legacy Five, board risk, post-quantum cryptography, AI cyber threats, SOCI Act, technology lifecycle, cost of downtime, CISO, digital resilience

KB sits down with Quinton Anderson, founder of Aigentsphere, to challenge one of the most repeated lines in AI governance: that a human in the loop keeps you safe. His view is that asking people to approve task after task just trains them to say yes.

They get into why you should never trust an agent but can trust the system around it, why an agent can’t be fully tested before it goes live, and why the brakes are what let you go fast. A sharp look at what it actually takes to govern agents at scale.

About Quinton:

Keywords: agentic AI, AI governance, human in the loop, human oversight, zero trust for agents, sociotechnical systems, continuous compliance, board accountability, AI safety, enterprise AI, Quinton Anderson, Aigentsphere, KBKast

Karissa Breen sits down with Gijo Varghese, Chief Security Officer at OT cyber security firm Secolve, to unpack an uncomfortable trade-off: the same connectivity and AI making power, water and transport smarter are also making them easier to break. Gijo explains how IT and OT convergence has widened the attack surface, why decades-old control systems were never built to touch the internet, and how a single IT intrusion can spill into the physical world. He walks through the incidents that prove it, from the 2015 Ukraine grid attack to Colonial Pipeline, where operators went to run the system by hand and found the people who knew how had all retired, to the Jaguar Land Rover breach that rippled through 5,000 suppliers and cost the UK economy billions. The throughline for boards and executives: a cyber incident stops being a security event the moment it becomes a public safety failure. Gijo makes the case for the kill switch, tested manual fallbacks, and treating resilience rather than compliance as the real measure of readiness.

About Gijo:

Secolve is Australia’s leading OT cybersecurity firm, providing cyber advisory, offensive security, and training services to mines, factories, hospitals, transport networks, and energy ecosystems. As Secolve’s first CSO, Gijo leads the firm’s consultancy and professional services team, transforming complex OT cyber risks into practical action across executive, engineering, and operational teams. 

Keywords: critical infrastructure security, OT security, IT/OT convergence, SCADA, ICS, cyber resilience, kill switch, Colonial Pipeline, Jaguar Land Rover, SOCI Act, CI45, incident response, operational technology, AI cybersecurity, public safety, board governance, cyber warfare, business continuity