'You don't win in security with a PowerPoint' At its TechSummit in Alpbach, Austria, Barracuda frames cybersecurity as a Formula 1 race. Every millisecond counts when powerful technology falls into the hands of hackers. Adam Khan, VP Global Security Operations & AI Security at Barracuda, opens his keynote with a striking comparison. "In Formula 1, you don't win just because you have the fastest car or the best driver, but because of a team that makes decisions every millisecond. The same happens in cybersecurity: attacks move faster and faster from one point to another. The combination of teams, tools, data, and AI helps you win the race." The comparison of security to a race will be frequently mentioned during the Barracuda TechSummit in the Austrian Alps. AI threatens to become the deciding factor in that race. Now that powerful models are being used by both the attacking and defending camps, it is a matter of who uses them the fastest, smartest, and most efficiently. History repeats itself. Klaus Gheri, VP Software Engineering, takes a step back in time. "In the past, we have often seen major changes that required architectural shifts. With AI agents, we are seeing that happen again. Everything you took for granted can suddenly be gone. The new normal today is that AI is already in your network." Khan sees what is happening with 'rogue' agents as a repetition of history and takes a forgiving stance toward the major AI players. "There have always been growing pains with new technology. That was true in the early days of the internet, when many data breaches occurred because everyone went online, and the same happened during the rise of the cloud. It never ultimately stopped us, but sometimes you have to go through that difficult phase first." One constant throughout that history is email as an attack vector. Phishing emails may be as old as the internet, but they remain a proven recipe for attackers. "If you look at them quickly, everything looks fine. It only starts once you open the file. AI can help in analyzing malicious emails and files," Khan adds. There have been growing pains with every new technology. This was also the case in the early days of the internet and the cloud. Adam Khan, VP Global Security Operations & AI Security Barracuda Identity as a new vector. This does not change the fact that Barracuda also sees new risks emerging, often related to AI. "Shadow AI goes much broader than IT. People put confidential info into AI systems, and organizations often don't know who is using which AI and how," says Brian Downey, head of product development. "Discovering this is the first step. Only then can you build a policy around it and protect data," his colleague Chris King assists him on stage. During the event, Barracuda announces AI Data Security to help organizations map out AI applications within their organization. Downey and King point to identity as the new battlefield in security. "This is the easiest thing for attackers to exploit. With the right keys, attackers walk right through the front door, and people do leave their keys behind sometimes," says Downey. "But 'non-human' identities pose just as much of a risk," King adds. To better protect digital identities, Barracuda acquired Evo Security, but the integration is still in progress. For Raynaud Schokkenbroek, Solutions Architect for the West European market, identity management should be at the top of every security leader's list. "Identity is the heart of access to your data, processes, billing, to everything. It is one of the most targeted elements. Privilege management is closely linked to that. If someone does get in, limit their ability to move freely." Who decides? Barracuda believes that the use of AI is necessary to stay one step ahead of attackers. Khan: "The time to intervene is shortening. Humans alone can no longer keep up with the volume and complexity of attacks. Attackers don't use closed models, but tools that are widely available, such as Copilot and ChatGPT. It is inevitable that we as defenders must also use AI." ITdaily ask Khan how he sees the division of labor between human security experts and AI agents. "We have already made significant progress in automating response. AI can follow up on alerts and reach a determination much faster than a human. AI is good at analysis; humans are good at the logic and the governance around it." The governance aspect is crucial for Khan. "Speed requires guardrails, otherwise a crash is inevitable. Without guardrails, you trust something without knowing what you are trusting. In the past, you could ask an AI model the same question ten times and get ten different answers. Now models are more reliable, but critical thinking and human intuition remain important to connect the dots. That balance is important." Schokkenbroek adds an extra criterion from the implementation side: "it must not be disruptive. For things that are by definition a problem, automation can work, and a human then decides if it falls outside the usual cases. The human-in-the-loop helps the model to learn. But we should not automate where it benefits no one." Invisible vulnerabilities. After speed, visibility is one of the biggest challenges for security today, according to Barracuda CISO Arve Kjoelen. "If we can bring together threats, exposure, assets, and controls, our challenges become visible. That is precisely the problem: assets come and go in the cloud, which also makes vulnerabilities dynamic. Visibility disappears." Kjoelen illustrates with an example. The number of known CVEs increases year over year, but according to him, only the tip of the iceberg is known. "We simply don't know how many vulnerabilities are in software. Researchers estimate that only ten, at most twenty percent, are known, and yet we continue to roll out new software." "With the use of LLMs, that could increase to sixty or seventy percent. LLMs discover many more critical vulnerabilities than humans. But then you also have to be able to patch those vulnerabilities. Remediation must therefore also automate. As a sector, we could do more with that 'boring' scientific work," Kjoelen adds. Behind every attack is a company. Barracuda's work will be far from over after the TechSummit. Or as Khan puts it: "You don't win the race with a PowerPoint". Those words sound rather ironic during a conference where presentations follow one another quickly, and where Khan himself juggles slides full of figures. "But behind every figure and attack is a person or company that may or may not have had a bad day. We as defenders must never forget that". Khan's biggest concern is when critical infrastructure is threatened. "We take things like electricity for granted. Society can only withstand a disruption for a short time before the fabric breaks. We must remain constantly alert". It is the logical conclusion of his own closing sentence. Behind every stopped attack is a person or a company that was not hit. That is what a security strategy must do, and that only happens in the field, not on slides. Behind every attack is a company that may or may not have had a bad day. Adam Khan, VP Global Security Operations & AI Barracuda

Barracuda launches industry-first AI Security and governance solution to accelerate Safe AI Adoption for Resource-Constrained Businesses and MSPs. GUEST RESEARCH: New Research Reveals Nearly Half of Senior IT Leaders Say Their Teams Lack the Skills to Securely Govern AI Already in Use, Creating Growing Security and Compliance Risks News highlights: * Barracuda launches Barracuda AI Data Security, an industry-first AI security and governance solution purpose-built for resource-constrained businesses and managed service providers (MSPs). * Barracuda AI Data Security is a complete, intelligent and easy-to-deploy solution that enables businesses to safely accelerate AI adoption while maintaining control of data security and compliance. * Addresses market needs as new research shows that nearly half of senior IT leaders say their teams lack the skills to securely govern AI already in use across their businesses. * Barracuda AI Data Security expands Barracuda's AI Security portfolio, laying out the foundation for upcoming innovations. Barracuda Networks, Inc., the cyber resilience platform leader making security easy for all size business, today launched Barracuda(R) AI Data Security, the industry's first AI security and governance solution purpose-built for resource-constrained organisations and managed service providers (MSPs). The complete, intelligent and easy-to-deploy solution enables businesses to confidently accelerate AI adoption by protecting sensitive data, enforcing responsible AI use and demonstrating compliance. Barracuda AI Data Security represents a major milestone in Barracuda's expanding AI Security portfolio. Built on the BarracudaONE(R) platform, the solution combines AI visibility, data protection, threat defense, and governance into a single integrated experience that makes AI security accessible and scalable. Businesses are embracing AI faster than they can securely govern it. New findings from Barracuda Research reveal that nearly half of senior IT leaders say their teams lack the skills required to securely govern AI already in use across their businesses. The consequences are already emerging: according to IBM's Cost of a Data Breach Report 2026, 68% of organisations that experienced a breach lacked policies to manage AI or detect shadow AI, highlighting the urgent need for stronger AI governance, protection and data security practices. "AI adoption is accelerating faster than most businesses can keep pace, creating one of the most urgent new security and compliance challenges organisations face today," said Neal Bradbury, Chief Product Officer at Barracuda. "Organisations need a practical way to secure AI as a new attack surface, protect sensitive data, enforce responsible use, and demonstrate compliance. At the same time, they need the confidence to embrace and accelerate AI adoption as a driver of productivity and innovation. Barracuda AI Data Security delivers those capabilities in a single, easy-to-deploy solution. This is the first of several innovations that will expand our AI Security suite and help businesses unlock AI's full potential, securely." Watch Bradbury's video message for deeper insight into today's announcement: https://www.youtube.com/watch?v=M-GMl2OGy7E A Complete, Intelligent Solution for Safe AI Adoption Barracuda AI Data Security helps businesses confidently embrace AI by combining security, governance and compliance into a single solution. It provides full visibility into AI usage, enables approved AI tools safely, inspects prompts and uploads in real time, and generates audit-ready evidence of policy enforcement. The result is faster AI adoption with stronger security, improved compliance and lower operational overhead. Powered by the Barracuda IQ(TM) AI engine, the solution extends protection across more than 1,300 generative AI (GenAI) tools, including Copilot, ChatGPT, Claude, and Gemini. It provides a fully integrated approach to AI security and governance: * AI data protection: Automatically blocks sensitive information such as customer data, passwords, financial records, and proprietary code from reaching GenAI services, with no initial configuration required. * AI threat detection: Blocks prompt injection, jailbreak attempts, hate speech, and policy violating topics using the Barracuda IQ AI engine's multilayered detection aligned to the OWASP LLM Top 10. * Compliance ready audit trail: Captures every AI interaction with one-click exports for insurers and auditors. * Flexible AI policies: Provides tailored rules for teams, tools and data types so businesses can safely say "yes" to AI where it adds value and "no" where it does not. Purpose Built for Resource-Constrained Businesses and MSPs Most AI security and governance offerings are designed for large enterprises, leaving smaller businesses and MSPs struggling with unnecessary complexity, specialised skill requirements and prohibitive costs. Barracuda AI Data Security is uniquely engineered for the realities of these underserved businesses, delivering enterprise-grade AI protection and governance through a streamlined, easy-to-manage experience that enables rapid, low-friction AI adoption. Barracuda AI Data Security is easily deployed in minutes with immediate protection through prebuilt sensitive data detection. It uses familiar firewall style workflows that eliminate the need for specialised AI security expertise and offers guided configuration through Barracuda's Bailey(TM) AI assistant. MSPs also gain centralised multitenant management and bundled pricing that fits naturally into their service models, enabling them to lead customers through responsible AI adoption with confidence. Availability Barracuda AI Data Security will be available in October 2026. It will also be included in Barracuda SecureEdge Premium Access at no additional cost. Strong Market Demand "Harnessing the benefits of AI in a safe, responsible and impactful way is a strategic priority for Horizon, but governance and oversight can be challenging and quite complex," said Jake Haacker, Chief Information and Security Officer at Horizon Health Alliance. "We need solutions that make protecting patient and other sensitive data seamless, while ensuring consistent policy enforcement and providing clear guardrails for responsible use. Barracuda AI Data Security will enable us to adopt AI faster and with far less risk, so our teams can stay focused on delivering exceptional care." "Our customers are already adopting AI and working with us to build AI policies that fit their specific security frameworks with a mindset of quick adoption. But policy only goes so far - both the business and IT need safeguards. We need controls that are simple, scalable and built for real-world environments," said Geoff Turner, Chief Executive Officer at Elevate Technology Group. "Barracuda AI Data Security gives us, as an MSP, a powerful way to help our customers embrace AI securely - with centralised management, strong data protection and clear evidence of compliance. It will enable us to lead customers through responsible and secure AI adoption with peace of mind and real controls protecting their business." "AI is reshaping the future of how our customers work, and MSPs have a critical role in helping them adopt it responsibly and at scale," said Adam Galfskiy, Head of AI at Aura Technologies. "What stands out about Barracuda AI Data Security is the forward-thinking foundation it gives us - a way to guide customers into an AI-driven future with clarity, control and confidence. The blend of visibility, governance and simplicity will empower organisations to innovate boldly without introducing new risk. It's a pivotal advancement for MSPs and the customers we serve." About Barracuda Barracuda is a leading global cybersecurity company delivering complete resilience made easy to buy, deploy and use through a partner-first model. Our intelligent BarracudaONE platform provides cyber resilience across email, data, applications, networks, and managed XDR within an open ecosystem. Trusted by hundreds of thousands of organisations and partners worldwide, Barracuda delivers industry-leading solutions and support, powered by people and enhanced by AI, for all size business.