AMD and F-Secure Tackle AI Risks with Adaptive Routing

AMD and F-Secure are joining forces to address growing cybersecurity challenges in the era of autonomous AI agents. By developing adaptive AI routing systems, the companies aim to secure digital journeys from devices to the cloud, ensuring privacy, safety, and efficiency as AI agents perform tasks on behalf of users. AI's role in digital interactions is evolving rapidly. Today's systems not only generate responses but also make decisions—comparing products, managing subscriptions, or even completing purchases. This shift means AI agents now interact with sensitive user data and potentially malicious third-party content, raising new risks around privacy, manipulation, and unsafe decision-making. F-Secure, a cybersecurity leader with 38 years of experience, has built TrustPath, an AI-native security platform designed to protect users throughout their digital journeys. Meanwhile, AMD is contributing its AI hardware and infrastructure expertise, leveraging solutions like AMD Ryzen™ processors for local AI processing and AMD EPYC™ and Instinct™ platforms for more complex cloud-based tasks. Together, the companies are looking to enhance trust in AI systems by ensuring routing decisions are tailored to the sensitivity and context of each interaction. In agentic browsing, AI systems assess content across various sources, including product reviews and third-party snippets. However, not all content is trustworthy. Malicious actors can use techniques like prompt injection to mislead the AI. This means routing decisions—whether an AI task should be processed locally, privately, or via external APIs—are critical. AMD and F-Secure's approach emphasizes "event-level handling," where trust is granted based on specific actions rather than entire domains. For example, a low-risk product comparison may be processed locally, while high-value transactions like payments or handling sensitive data would require private infrastructure with stricter security controls. The collaboration between AMD and F-Secure could set new standards for AI application development. Key benefits for developers include: AMD's open-source Lemonade platform simplifies deployment by allowing developers to integrate local AI models seamlessly with AMD hardware. For enterprises, this architecture offers flexibility to adapt to evolving threats, regulations, and user needs without requiring constant application-level changes. AMD's acquisition of Silo AI in 2024, Europe's largest private AI lab, has been pivotal in expanding its AI capabilities. Silo AI's expertise has enabled AMD to develop sovereign AI solutions, including multilingual language models and advanced inference capabilities on AMD Instinct GPUs. F-Secure's connection to this ecosystem, through its collaboration with Qutwo, underscores a shared commitment to building AI-native consumer protection tools. This partnership also aligns with a growing market focus on AI security. As of September 8, 2026, AMD's stock price was $477.57, up 4.69% over the past 24 hours, reflecting investor confidence in its AI-forward strategy. Meanwhile, F-Secure's TrustPath platform continues to position the company as a leader in AI-era cybersecurity. As AI systems become more autonomous, securing digital journeys will be critical for user trust. By combining AMD's AI hardware and routing capabilities with F-Secure's cybersecurity expertise, this collaboration aims to create a scalable, privacy-focused framework for the agentic era. Developers and enterprises alike should watch for advancements in dynamic AI routing as the partnership evolves.