Daily Cyber & AI Briefing with Michael Housch. This episode was published automatically and includes the assembled audio plus full transcript.
Today’s cyber and AI risk environment is defined by a relentless pace of attacks, a surge in exploitation of zero-day vulnerabilities, and a growing sense of urgency around AI governance. We’re seeing adversaries target the very infrastructure designed to keep organizations safe, while AI is increasingly being used both as a tool for defenders and as a weapon for attackers. The stakes are rising, and the gap between risk and effective response is becoming more apparent—especially in critical infrastructure and enterprise environments.
Let’s start with the most pressing developments on the cyber front. Over the past 24 hours, attackers have been actively exploiting vulnerabilities in widely deployed security appliances and software. This includes Fortinet, Palo Alto, Ivanti, Cisco, and even WinRAR. These aren’t obscure products—they’re foundational elements in many organizations’ security stacks, and that’s exactly why they’re such attractive targets.
A particularly concerning campaign involves hackers exploiting Fortinet vulnerabilities to deploy a stealthy malware strain called PEERSTITCH on FortiGate firewalls. The sophistication here is notable: PEERSTITCH is engineered for stealth and persistence, allowing attackers to maintain long-term, undetected access to compromised firewalls. This isn’t just about breaking in—it’s about staying in, quietly monitoring traffic, and waiting for the right moment to move laterally across the network. The fact that attackers are focusing on security infrastructure itself is a wake-up call. These devices are often trusted implicitly, and if they’re compromised, the attackers can potentially pivot deeper into your environment without raising alarms.
For security leaders, this underscores a few critical imperatives. First, timely patching is non-negotiable. If you’re running FortiGate firewalls, you need to be aware of the latest advisories and ensure you’re not lagging behind on updates. Second, continuous monitoring of your security appliances is essential. Don’t assume that just because a device is labeled “security,” it’s immune from attack. In fact, these devices should be among the most closely watched assets in your environment. Finally, any anomalous behavior observed on these devices should trigger rapid investigation. Attackers are counting on organizations being slow to respond or failing to notice subtle changes.
Moving to Ivanti, CISA has issued a fresh alert regarding active exploitation of a zero-day vulnerability in Ivanti Connect Secure VPN appliances. Attackers are leveraging this flaw to steal credentials and sensitive data, posing a direct threat to organizations that rely on Ivanti for remote access. The practical takeaway here is clear: apply available mitigations immediately, monitor for signs of compromise, and review your remote access policies. VPN appliances are high-value targets—if compromised, they can give attackers a direct line into your internal network, bypassing many traditional defenses.
It’s a similar story with Palo Alto Networks’ PAN-OS. A critical zero-day vulnerability, designated CVE-2026-0322, is being actively exploited in the wild. Organizations using affected versions are urged to patch immediately. The risk here isn’t hypothetical—attackers can use this flaw to gain unauthorized access or disrupt security operations. This incident reinforces the importance of rapid vulnerability management and close coordination with vendors. When a zero-day hits, every hour counts. Make sure your team is plugged into vendor threat intelligence and patch guidance.
Cisco is also in the spotlight, having issued an urgent advisory for customers to patch a critical vulnerability in its email gateway products. Email gateways are a prime target for attackers looking to intercept or manipulate business communications. If you’re running Cisco gateways, prioritize patching, review your email logs for any suspicious activity, and enhance monitoring. Email remains one of the most common initial access vectors for attackers, and a compromised gateway can have broad implications for data loss and business continuity.
We’re also seeing attackers exploit a critical authentication bypass vulnerability in Fortinet FortiClient EMS. What’s especially concerning here is that a proof-of-concept exploit has already been released. This lowers the barrier to entry for would-be attackers, increasing the risk of widespread exploitation. If you haven’t already, apply patches immediately and monitor for unusual access patterns. Authentication bypass vulnerabilities are among the most dangerous, as they can allow attackers to walk right in without needing valid credentials.
Legacy software isn’t immune, either. A threat group identified as UAT-6275 is exploiting a vulnerability in WinRAR to target government organizations in the Commonwealth of Independent States. This campaign is a reminder that unpatched, widely used utilities can present a significant risk, especially in government and critical infrastructure environments. Security leaders should ensure all endpoints are updated and consider restricting the use of high-risk utilities like WinRAR, especially if they’re not strictly necessary for business operations.
Let’s shift gears to the intersection of cyber and AI risk—a space that’s evolving rapidly and presenting new challenges for organizations. A recent report found that AI is now used in 84% of ransomware attacks. That’s a staggering statistic. Adversaries are leveraging AI to automate targeting, evade detection, and compromise identity infrastructure. This trend amplifies the risk to organizations’ authentication and access systems. It’s not just about defending against brute-force attacks anymore; it’s about defending against intelligent, adaptive adversaries who can move faster and more quietly than ever before.
The practical implication is that organizations need to enhance their identity protection strategies. Multi-factor authentication is a must, but it’s not sufficient on its own. AI-driven anomaly detection should be part of your toolkit. Monitor for unusual login patterns, privilege escalations, and other indicators of compromise. And don’t overlook the human element—ensure that employees understand the risks and know how to recognize and report suspicious activity.
Ransomware actors are also escalating their tactics. We’re seeing a shift in focus—not just targeting organizations, but also going after employees and their families. This is a significant evolution in ransomware strategy. By using social engineering and extortion, criminals are increasing the pressure on organizations to pay up. This raises the risk of insider threats and can cause significant reputational damage. Organizations need to bolster employee awareness and support, and ensure that incident response protocols are up to date. It’s not just about protecting data and systems anymore—it’s about protecting people.
The convergence of AI and cyber threats is also putting identity infrastructure under sustained attack. As attackers weaponize AI, traditional defenses are being bypassed with increasing frequency. This demands a reassessment of identity and access management controls. Organizations need to move beyond static defenses and invest in AI-driven detection and response capabilities. The goal is to spot and stop attacks before they can do real damage.
Now, let’s talk about the broader AI landscape. Enterprises are entering a new phase of AI adoption—one that’s focused on risk management. CompTIA research indicates that organizations are moving from experimentation to a more mature, risk-aware approach. Security concerns are now a top priority, and there’s a growing recognition of the need for formal governance, risk assessment, and controls around AI deployments.
But there’s a gap. An IBM study found that 80% of UK enterprises now escalate AI-related incidents to their boards, but many report lagging governance and oversight. This highlights a growing materiality of AI risks at the executive level, but also a lack of mature governance structures and incident response plans for AI systems. The trust gap in enterprise AI is real. Stakeholders are concerned about transparency, bias, and security. The recommended solution is robust governance—continuous monitoring, clear accountability, and alignment with regulatory expectations.
For CISOs and risk executives, the imperative is clear: accelerate the development of AI governance frameworks that go beyond static compliance. Static compliance approaches—where you check the box once and move on—are insufficient, especially in critical infrastructure. Instead, what’s needed is continuous monitoring and adaptive controls. AI systems are dynamic by nature; the risks they present can change over time, and your governance approach needs to reflect that.
Expert commentary stresses that for critical infrastructure, continuous monitoring and real-time risk detection are necessary to manage evolving threats and maintain resilience. CISOs should be advocating for dynamic governance models and investing in the tools and processes that enable ongoing oversight of AI-enabled systems. This isn’t just about regulatory compliance—it’s about operational resilience and maintaining trust with stakeholders.
Let’s distill the strategic implications from today’s developments. First, the wave of zero-day vulnerabilities in security infrastructure means organizations need to accelerate patching and implement continuous monitoring. Attackers are increasingly targeting the very tools meant to protect us, and complacency is not an option.
Second, the weaponization of AI in ransomware an