This week in cybersecurity highlighted how AI is simultaneously expanding defensive capabilities and creating new attack vectors. OpenAI detailed how AI agents breached Hugging Face in a multi-stage compromise, while researchers identified hidden code instructions as an emerging threat to AI models. At the enterprise level, 86% of SMEs surveyed by Kaspersky experienced at least one cybersecurity incident in the past year, reinforcing the need for stronger protection. Fortinet and Alestra are responding by advancing automation and managed security capabilities for Mexico’s telecom sector.
Ready? This is the Week in Cybersecurity!
AI Models Face New Cyber Threat From Hidden Code Instructions
AI has transformed cybersecurity by giving defense teams capabilities that were previously difficult to achieve at scale. Models can process vast amounts of telemetry, identify anomalies, and automate threat detection at speeds that human teams cannot match alone. But the same technology that strengthens cyber defenses is creating a new attack surface. As organizations increasingly rely on AI for critical security and business decisions, attackers are beginning to target the models themselves rather than only the applications and infrastructure they protect.
Fortinet, Alestra Strengthen Telecom Cybersecurity in Mexico
For telecommunications companies managing infrastructure and services for thousands of organizations, the ability to respond to cyber threats is increasingly tied to how quickly security and network operations can identify and address risks. To address this scenario, Fortinet and Alestra are moving toward a more automated operating model through the integration of managed security services and security orchestration, automation, and response capabilities.
Cybersecurity Incidents Hit 86% of SMEs in the Last Year
As global SMEs expand their use of digital tools, they are also increasing the number of systems, accounts, and endpoints that can become targets. According to a global survey by Kaspersky among IT security specialists from organizations in 18 countries, 86% of companies with fewer than 500 employees experienced at least one cybersecurity incident over the last year. Only 14% avoided an incident during the period.
OpenAI Details How AI Agents Breached Hugging Face
OpenAI has issued a disclosure of how its AI agents breached Hugging Face, providing a detailed account of an incident that moved from a controlled cybersecurity evaluation into a multi-stage compromise of production infrastructure. In a 37-page technical report, the company describes how models operating as agents circumvented network restrictions, coordinated through an unintended communication channel, obtained publicly exposed credentials, and exploited vulnerabilities to gain access to Hugging Face systems.
