
Sign up to save your podcasts
Or


Connect With Us
AI adoption has become the defining challenge of modern cybersecurity programs, and the hurdle most organizations face is not the technology itself. It is the skill, the knowledge, and the trust required to deploy it responsibly. Michael Dawisha joins Paul Marco for a wide-ranging, career-spanning conversation about AI adoption, the generational shift in how technology is understood, the evolution of the CIO role, and the growing question of what humanity is willing to trade for productivity.
The episode centers on a pointed finding from the 2026 World Economic Forum Global Cybersecurity Outlook: 54% of organizations cite insufficient knowledge or skill as the primary hurdle to adopting AI for cybersecurity. Michael, a 30-year CIO with experience spanning mental health, higher education at Michigan State University, and now Genesee County government, brings a practitioner’s eye to that number. Paul reframes it against the practical reality of asking an AI engine the right questions, interpreting its outputs, and knowing when to trust them.
You will discover why Michael believes the modern CIO must shed the techie-only identity of the past and earn a seat at the board table, the pivotal moment early in his career that set him on that path, and why the same lesson now applies to the CISO role. The conversation examines why using AI is no longer the cheating behavior, and why failing to use it is quickly becoming the real organizational liability. Michael walks through his team’s framing of AI as an intern, capable of producing strong first-pass work that still demands human review before it goes anywhere meaningful.
The discussion widens into the unsettling territory of self-modifying code, neural networks that program themselves, and the increasingly real possibility that frontier models are developing behaviors their own creators cannot fully explain. The panel references recent reporting on AI models leaving notes for their successors, examines what it means when the people building these systems openly admit they do not understand the emergent behavior, and draws parallels to the long path society traveled from fearing Google to embracing it, with the critical difference that AI is evolving at machine speed.
The conversation then turns to the human element, where the sweet pull of productivity consistently overrides caution. Michael and Paul explore why users resist guardrails when AI delivers results, why the next generation of workers enters the workforce with no memory of the privacy norms that shaped prior eras, and why homogenization of thought, tone, and output is a real cultural risk when everyone uses the same engine to produce the same content.
The episode closes on a sobering observation about the arc of human trade-offs. Society has repeatedly bargained freedom for safety, privacy for convenience, and attention for connection. The next bargain may be agency itself, as more people defer judgment, planning, and creativity to the model, trusting that the totality of human knowledge sitting at their fingertips surely knows better than they do.
Essential listening if you care about AI adoption, cybersecurity skills development, the evolving CIO and CISO role, generational shifts in technology culture, or the deeper question of what AI will cost us even as it gives us more. If you are responsible for protecting people, systems, or data, this conversation gives you both the practical perspective and the philosophical grounding to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
The adversary playing field has fundamentally expanded, and the organizations still operating under old assumptions about who targets whom are the ones most exposed. Raul Rosado joins Owahn Bazydlo and Paul Marco for a wide-ranging, higher-education-grounded conversation about nation-state cyber activity, the collapse of statecraft boundaries, insider threat through the hiring process, and why resilience has quietly overtaken prevention as the real measure of a cybersecurity program.
The episode centers on a striking finding from the 2026 Verizon Data Breach Investigations Report: Amazon disclosed thwarting more than 1,800 North Korean remote worker infiltration attempts in 2025, identified in part through a distinctive 110 millisecond keystroke input lag. Raul, Chief Information Officer for Universidad del Sagrado Corazón in San Juan, Puerto Rico, with a background spanning banking risk management, Deloitte consulting, and a CISO role at a startup, brings a practitioner’s eye to the number. The panel is clear that 1,800 attempts is almost certainly a conservative snapshot, and that similar activity is accelerating across every industry and institution type, not just hyperscale technology companies.
You will discover why North Korea pioneered the strategy of placing workers inside Western organizations as a sanctions-evasion revenue stream and intelligence pipeline, why this represents one of the hardest categories of insider threat to detect, and why the Amazon detection signal, a 110 millisecond keystroke lag, is itself a reminder that establishing a baseline of normal behavior is now a foundational control. The conversation also examines parallel activity from Iran and China, including recent reports of exfiltration from major technology firms, and why universities have become particularly attractive targets given their combination of valuable research, open collaboration culture, and distributed computing resources.
The discussion widens into a fundamental shift in how threat models must be constructed. Nation-state actors are no longer the exclusive concern of the Fortune 50 and the defense industrial base. Small organizations, mid-market firms, nonprofits, and educational institutions all sit inside interconnected digital supply chains that make them viable stepping stones toward higher-value targets. The panel examines why the long-standing defensive philosophy of being annoying enough to make attackers move on does not apply to mission-driven adversaries, who will persist until the objective is achieved regardless of friction.
The conversation also touches on the parallel accountability problem in the private sector, including the recent 24 billion privacy lawsuit against Meta, and the question of whether research ethics structures like Institutional Review Boards should extend beyond academia into any organization conducting large-scale experimentation on user data. Raul walks through how universities approach IRB oversight and what that framework might look like if applied to the technology industry.
The episode closes on a shared conviction that cybersecurity success is no longer measured by the number of attacks prevented, but by how effectively an organization operates when something goes wrong. Resilience, muscle memory, response discipline, and knowing your own environment better than the adversary ever can are the enduring foundations beneath every new wave of threat sophistication.
Essential listening if you care about nation-state cyber activity, insider threat, higher education cybersecurity, data ethics, or the evolving reality of defending interconnected environments against motivated adversaries. If you are responsible for protecting people, systems, or data, this conversation gives you both the realism and the operational perspective to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
The pace of change in cybersecurity has never been faster, and the forcing function is the same technology that is enabling the next generation of attackers. Evan Obranovic joins Owahn Bazydlo and Paul Marco for a wide-ranging, education-grounded conversation about AI-enabled threats, the collapse of the barrier to entry for cybercrime, insider risk in the student era, and the growing fatigue around AI being bolted onto every tool in the stack.
The episode works through two complementary findings. The first, from the 2026 Google Cloud Cybersecurity Forecast, is that December 2025 saw the discovery of Void Link, a malware framework written in six days by an AI agent, described as a point of no return for automated threat development. The second, from the 2026 World Economic Forum Global Cybersecurity Outlook, is that 94% of respondents identify AI as the most significant driver of cybersecurity change in 2026, and 87% flag AI-related vulnerabilities as the fastest growing cyber risk in 2025. Evan, an 18-year education veteran now serving as executive director of technology for a Michigan district that includes a data center supporting other districts and local organizations, brings a practitioner’s eye to both numbers.
You will discover why the economics of malware has been fundamentally rewritten, with specialized development teams no longer required to produce functional ransomware or exploitation frameworks. The group examines how unrestricted AI models have erased the traditional gating function around cybercrime, leaving only intent as the differentiator between a curious hobbyist and a capable adversary. The panel also walks through the real operational experience of discovering homegrown phishing campaigns inside a school district, where students who understand the tools faster than the adults around them are already pressure testing institutional defenses.
The conversation widens into the insider threat implications of putting agentic AI capability into every user’s hands, the ongoing challenge of walled-garden tools that remain susceptible to in-environment data exposure when prompted the right way, and the uncomfortable reality that most data protection assurances still rest on vendor terms and conditions rather than enforceable technical boundaries. Evan, Owahn, and Paul also address the growing fatigue with AI being pitched into every vendor conversation, and why mission-driven organizations should push back on tool adoption that does not serve a clearly defined problem.
The episode closes on a shared conviction that speed is the defining theme of this moment in cybersecurity. The organizations that survive this transition will be the ones that pair fast adoption with disciplined inventory, strong configuration standards, honest conversations with their users, and the willingness to slow down when the question of why, what, and who is making the decision has not been answered.
Essential listening if you care about AI-enabled threats, insider risk, education cybersecurity, vendor AI fatigue, or the operational reality of defending modern environments at machine speed. If you are responsible for protecting people, systems, or data, this conversation gives you both the urgency and the perspective to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
Ransomware remains the dominant breach pattern in cybersecurity, but the trajectory beneath that headline is more encouraging than most practitioners realize. Bipin Prakash joins Paul Marco for a wide-ranging, education-grounded conversation about ransomware economics, the shifting threat surface from email to browser, the sophistication of modern social engineering, and why the industry’s default pessimism may be masking real defensive progress.
The episode works through two complementary findings from the 2026 Verizon Data Breach Investigations Report. The first: ransomware appeared in 48% of all breaches in 2025, up from 44% the previous year, and in 77% of system intrusion pattern breaches. The second: the median ransom payment declined to $139,875 in 2025, down from $150,000 the prior year, continuing a two-year downward trend. Bipin, a technology director with two decades of experience in K-12 education, brings a practitioner’s eye to both numbers. Paul reframes his initial expectation that the ransomware prevalence figure would land in the 80s or 90s, and the panel coalesces around a surprisingly optimistic read: defenses are improving, resiliency is up, and attackers are being forced to accept smaller payouts.
You will discover why ransomware is inextricably linked to the rise of cryptocurrency, and why the industry’s two-decade shift from credit card fraud to direct extortion traces back to that single enabling technology. The group unpacks the modern ransomware ecosystem, where access brokers, exfiltration teams, and encryption operators function as separate specialized businesses, each taking a slice of a commoditized attack chain. The conversation then examines whether AI-enabled development, agentic swarms, and ransomware-as-a-service platforms are quietly driving per-incident payouts down through sheer volume and reduced overhead on the attacker side.
The discussion widens into the operational reality of defending modern environments, where the browser has replaced email as the fastest-growing data loss vector, where shadow AI usage has made the perimeter conversation obsolete, and where cloud workloads have outpaced the governance frameworks built for on-premise infrastructure. Bipin and Paul explore the sophistication of modern social engineering, including real-world examples of new hires receiving impersonation attempts within hours of starting, the aggregation of public data through sites like TruePeopleSearch, and the use of multimodal AI agents to execute convincing voice and video impersonation at scale.
The episode closes on the authentication paradox that defines our moment. Passwords gave way to two-factor, two-factor gave way to voice verification, voice gave way to video verification, and video is now giving way to family safe words. The industry has traveled a full circle back to shared secrets, which is itself a reminder that human verification, lived experience, and curiosity are the durable advantages no model can replicate.
Essential listening if you care about ransomware trends, the economics of extortion, browser-based data loss, shadow AI, or the evolving reality of social engineering in an AI-enabled world. If you are responsible for protecting people, systems, or data, this conversation gives you both the measured optimism and the operational perspective to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
Not every DLP alert tells the full story, and the ones that do are usually the tip of a much larger iceberg. Paul Zimmerman joins Owahn Bazydlo and Paul Marco for a wide-ranging, education-grounded conversation about AI governance, intellectual property risk, user behavior, and why the next phase of cybersecurity will be shaped as much by optimism and design as by fear.
The episode centers on a sharp finding from the 2026 Verizon Data Breach Investigations Report: in 3.2% of DLP policy violations, research and technical documentation was uploaded to unauthorized AI systems, creating direct intellectual property risk exposure. Paul Zimmerman, a technology director based in Hailey, Idaho, whose career began in Air Force encryption work and moved through MSP operations before landing in K-12 technology leadership, brings a practitioner’s eye to the number. The panel is clear that 3.2% is almost certainly a significant undercount, representing only what current DLP tooling can actually see, with the real exposure likely sitting several multiples higher.
You will discover why the threat model behind this statistic is far more complex than a single user pasting a single document into a free AI tool, and why the real exposure now extends through chains of subprocessors embedded in every modern SaaS platform. The group examines why FERPA, HIPAA, and other regulatory frameworks are already struggling to keep pace with frontier model capabilities, why organizations that fail to provide sanctioned AI tools inevitably drive their people toward unauthorized ones, and why transparency across the data processor chain has become one of the most important and least enforced controls in enterprise cybersecurity.
The conversation then widens into the broader AI governance question, drawing parallels to Google’s early days when similar outrage over data harvesting ultimately gave way to an entire economy built around responsible data use. Paul Zimmerman challenges the industry’s reflexive doom and gloom with a reminder that the automobile arrived in a world with no rules, and that human beings built the guardrails as they went. The group explores the collapse of traditional bug bounty programs under the weight of AI-generated spam submissions, the shift from expert-driven vulnerability discovery to tool-enabled participation, and the ongoing erosion of research spaces that once depended on expert curation.
The episode closes on a shared philosophy that Paul Zimmerman calls “freedom in a box.” Build the box through policy, technical controls, user training, and sanctioned AI tooling, then let people operate freely inside it. The organizations that embrace this model, that treat statistics as directional rather than definitive, and that pair governance with genuine curiosity about what AI can enable, will be the ones that turn today’s risk into tomorrow’s advantage.
Essential listening if you care about data loss prevention, AI governance, intellectual property protection, FERPA and HIPAA compliance, or the future of responsible AI adoption in regulated industries. If you are responsible for protecting people, systems, or data, this conversation gives you both the realism and the optimism to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
Twenty-seven seconds. That is the fastest e-crime breakout time on record in the 2026 CrowdStrike Global Threat Report, and it reframes almost every assumption we make about defensive response. Albert Galeas joins Owahn Bazydlo and Paul Marco for a candid, education-grounded conversation about cyber hygiene, governance, the limits of security tooling, and why the fundamentals still decide who wins when the attack clock is measured in seconds.
The episode centers on the CrowdStrike finding and what it actually reveals: that the fastest observed pivot from initial access to a second system in the environment took less than half a minute. Albert, a technology director whose career began in the classroom and moved into IT leadership during COVID, brings a practitioner’s eye to the number. Paul and Owahn draw the line from that measurement to the preconditions that make it possible, including flat networks, shared admin credentials, missing access segmentation, and unrotated built-in accounts. The panel is clear that 27 seconds is not a story about attacker genius, it is a story about defender hygiene.
You will discover why the CIS Group 1 Controls remain one of the most under-utilized defensive frameworks available to resource-constrained organizations, why security awareness training continues to deliver outsized returns even in environments with mature email filtering, and why asset inventory and governance are the real foundation beneath every other control. Albert walks through a seemingly small operational change, routing all new devices through a central configuration office before they ever touch the production network, as a reminder that defensive maturity is built from procedural discipline, not from product acquisition.
The conversation then widens into the near-term future, with the group examining how agentic AI is likely to compress breakout times further, why defenders will need their own agentic blue-team capability to keep pace, and why the current wave of AI-enabled attacks is likely to shift the industry’s focus from end-user compromise back toward automated vulnerability exploitation. Albert, Owahn, and Paul also return repeatedly to the attacker’s advantage: the adversary needs only one way in, while the defender is responsible for the entire perimeter.
The episode closes on a shared conviction that simplicity, discipline, and self-knowledge are the durable foundation beneath every new wave of tooling. Attackers will keep getting faster. The organizations that keep their house in order, that know their inventory, that patch consistently, that disable accounts on exit, and that make the environment annoying enough to drive the adversary elsewhere, will keep buying themselves the time their analysts need to respond.
Essential listening if you care about cyber hygiene, foundational controls, security awareness, agentic AI threats, or the future of defensive operations. If you are responsible for protecting people, systems, or data, this conversation gives you both the urgency and the perspective to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
There is finally a bright spot in the AI governance conversation, and it comes from an unexpected place: the companies themselves. Michael Brady joins Paul Marco for a candid, financial-services-grounded conversation about AI governance, shadow AI, third-party risk, and why defense remains one of the hardest jobs in cybersecurity.
The episode centers on an encouraging finding from the 2026 World Economic Forum Global Cybersecurity Outlook: 64% of organizations now have a process to assess the security of AI tools before deployment, nearly double the 37% reported in 2025. Michael, a CIO and CISO in the financial services industry with over 25 years of experience across verticals ranging from armored cars to satellites, brings a practitioner's eye to the number. Paul draws the line from written policy to operational reality, pressing on whether these processes are actually trained, executed, and enforced inside the business, or whether they are simply AI-generated artifacts sitting in a governance folder.
You will discover why Fannie Mae and Freddie Mac's AI governance letters are forcing the entire mortgage banking ecosystem to formalize its posture, why vendor renewals have become one of the most under-managed risks in financial services, and why cascading subprocessor terms and conditions are quietly transferring AI data exposure down a chain that most legal teams cannot keep up with. Michael walks through how healthcare underwriters that turned AI fully on and removed humans from the loop ended up making biased decisions that cost them their business, as a cautionary tale for any organization tempted to let AI replace human judgment in high-stakes workflows.
The conversation widens into the cloud parallel, with Michael and Paul tracing how AI adoption is moving at a velocity that makes early cloud adoption look slow, and why that velocity is itself a risk factor. The group examines why shadow AI, personal accounts, and fly-by-night wrapper tools remain a major blind spot even in mature programs, why sales teams are being overwhelmed by shiny-object technology bundles, and why prompt injection and model manipulation are only the first wave of a longer attack surface problem. Michael also makes a strong case for managed services and global threat visibility, arguing that no single organization can staff its way through the current pace of change.
Michael and Paul close on the deeper reality underneath all of it: the asymmetry of cyber defense, the invisibility of successful protection, and the budget conversations that follow when nothing goes wrong. It is a practical reminder that cybersecurity leaders are judged by outcomes they can rarely show, and that AI will not change the fundamental math of defending complex, human-built environments.
Essential listening if you care about AI governance, shadow AI, third-party and subprocessor risk, financial services cybersecurity, or the future of defense in a machine-accelerated world. If you are responsible for protecting people, systems, or data, this conversation gives you both the optimism and the operational urgency to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
Shadow IT never went away, it just learned to speak in prompts. Lou Mazzucchelli joins Owahn Bazydlo and Paul Marco for a wide-ranging, historically grounded conversation about AI governance, corporate culture, data as an asset, and why the newest problems in cybersecurity keep rhyming with the oldest.
The episode centers on a striking finding from the 2026 Verizon Data Breach Investigations Report: 67% of users access AI services via non-corporate accounts on corporate devices, and 45% of employees are now considered regular AI users on corporate devices, up from 15% the previous year. Lou, who holds the first undergraduate degree in AI ever granted in the United States and whose career spans everything from IBM 1401 mainframes to venture capital to building a corporate culture with such staying power that 180 former employees held a reunion 10 years after the company was gone, turns that number into practical reality. Owahn and Paul draw the line from personal productivity pressure to enterprise data leakage.
You'll discover why unauthorized AI use is not a novel threat but a repeat of decades of shadow IT behavior, why organizations that fail to provide sanctioned tools drive their people to rogue platforms, and why corporate data assets should appear on the balance sheet as seriously as gold bars. Lou walks through the Maersk incident, where a single server offline in Ghana became the seed of a global recovery, as a reminder of what enterprise data is truly worth when it is gone. The group also examines why standard vendor sourcing practices, if honestly applied, would disqualify most of today's AI providers on financial viability alone, and why corporate FOMO is overriding that discipline at the board level.
Lou, Owahn, and Paul then dig into the deeper shift underneath all of it: the enduring importance of culture, intentional hiring, quiet workspaces, and the small signals of respect, like a $12 mouse, that determine whether employees pick up the phone at one in the morning. The conversation reframes shadow AI as a human and organizational problem rather than a technology problem, and closes with a practical call to use existing telemetry, long-tail web log analysis, and honest risk analysis to surface what employees are actually doing before the data walks out the door.
Essential listening if you care about AI governance, shadow IT, third-party risk, data valuation, or the future of corporate culture in a machine-accelerated world. If you are responsible for protecting people, systems, or data, this conversation gives you both the historical perspective and the operational urgency to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
AI is accelerating attacks faster than most defenders can adapt, but the surprise is how little of that acceleration is actually new. Dustin Brown joins Owahn Bazydlo and Paul Marco for a wide-ranging conversation about agentic AI, third-party risk, critical infrastructure, and why the fundamentals still decide who wins.
The episode opens with a striking finding from the 2026 Verizon Data Breach Investigations Report: less than 2.5% of AI-assisted malware observations involve rare techniques with one or fewer known malware examples, indicating that AI is currently accelerating known attack methods rather than unlocking novel ones. The panel turns that number into practical reality. Dustin, a Texas-based technology leader working across IoT and critical infrastructure, brings a defender's eye to the space where the physical and digital layers meet, while Owahn and Paul draw the line from commodity exploits to enterprise exposure.
You'll discover why agent-driven commerce is reshaping the risk surface, with an estimated 10% of transactions already originating from agents and projections pointing toward 90% within a year. The group unpacks why third-party and fourth-party risk questionnaires have become an operational burden on small and mid-market firms, why trust portals are emerging as a smarter path to vendor transparency, and why the SOC 2 attestation process forces the kind of vendor discipline most organizations delay far too long.
Paul and Owahn then widen the lens into the reality of what Dustin calls vibe-coded "plastic apps," the security debt they generate, and why AI functions as a force multiplier that rewards skilled hands and punishes unskilled ones. The conversation examines how attackers are simply running old playbooks at machine speed and machine volume, why obscurity as security no longer holds, and how observability, asset knowledge, and disciplined patching remain the durable foundation beneath every new wave of tooling.
Dustin, Owahn, and Paul also dig into the deeper shift underneath all of it: the human trade of agency for convenience, the cultural gap in AI literacy, and the growing need for personal security awareness at the same level enterprises have long demanded. It is a practical reminder that cybersecurity outcomes still trace back to intent, discipline, and preparation, no matter how capable the tools become.
Essential listening if you care about AI-enabled threats, third-party risk, critical infrastructure, or the future of agentic systems inside the business. If you are responsible for protecting people, systems, or data, this conversation gives you both the urgency and the perspective to think more clearly about what comes next.
https://www.talas.io/podcast
Connect With Us
In this episode of Cybersecurity Perspectives, Paul Marco and Owahn Bazydlo unpack a striking data point from the 2026 Verizon Data Breach Investigations Report: the median threat actor researched or used AI assistance across 15 different MITRE ATT&CK techniques in 2025, with some actors leveraging as many as 40 to 50. What follows is a candid discussion of what happens when the very frameworks defenders rely on become the roadmaps adversaries feed to unrestricted large language models.
The conversation frames a fundamental asymmetry that has always existed in cybersecurity but is now accelerating at machine speed. MITRE ATT&CK was built to codify known adversary tradecraft and give defenders a common language for identifying coverage gaps. That same catalog, once handed to an AI agent, becomes an efficient checklist for probing enterprise environments. Attackers need only one technique to succeed. Defenders need every control to hold. AI compresses the time and expertise required on the offensive side, while defenders remain constrained by procurement cycles, architectural complexity, incomplete asset inventories, and log coverage gaps that no framework alignment exercise can quickly resolve.
The episode also touches on broader industry context, including recent departures from frontier AI labs by researchers raising alignment concerns, and updated ransomware payment data pointing to manufacturing, education, and healthcare as the most frequently paying verticals. Together, these threads reinforce the central takeaway: the defender's advantage now depends less on tooling sophistication and more on operational discipline, environmental clarity, and the ability to contain a compromise before it outpaces the humans defending against it.
https://www.talas.io/podcast