{"id":46605,"date":"2025-07-07T07:10:50","date_gmt":"2025-07-07T07:10:50","guid":{"rendered":"https:\/\/www.mon-agent-ia.fr\/blog\/?p=46605"},"modified":"2025-07-07T07:10:51","modified_gmt":"2025-07-07T07:10:51","slug":"ai-agent-the-new-discreet-cybersecurity-threat-for-businesses","status":"publish","type":"post","link":"https:\/\/www.mon-agent-ia.fr\/blog\/en\/ai-agent-the-new-discreet-cybersecurity-threat-for-businesses\/","title":{"rendered":"AI Agent: The New, Discreet Cybersecurity Threat for Businesses?"},"content":{"rendered":"

As we approach 2025, businesses have never faced more cybersecurity challenges, particularly with the emergence of artificial intelligence (AI) agents. These promising tools, which simplify many processes, can also become vectors for unsuspected threats. The SailPoint report, released in July 2025, highlights a worrying reality: the growing adoption of AI agents is accompanied by new and complex risks. While 82% of companies are adopting these technologies, a majority feel vulnerable to unexpected and sometimes harmful behaviors. In this context, how can companies navigate this landscape of potential threats?<\/p>\n\n

Artificial intelligence agents: double-edged allies<\/h2>\n\n

AI agents, such as those developed by companies such as Thales and Orange CyberDefense, are integrated into data management and analysis solutions. While they offer increased efficiency, these tools also open the door to a multitude of security issues, as their access is often far greater than that of traditional employees. Expanded Access and Amplified Risks<\/strong> The SailPoint report highlights that 54% of IT professionals believe that AI agents have access to many more systems than traditional users. Some of the main risks identified include: Accessing sensitive data without proper justification (60%).<\/strong>Performing unintended tasks that can lead to errors (58%).<\/p>\n\n

Sharing confidential information without authorization (57%).<\/h3>\n\n

Making decisions based on unreliable data (55%).<\/p>\n\n

  • Disturbing Incidents: The Reality on the Ground<\/li>
  • Companies are facing alarming friction scenarios. Nearly 23% of them have already experienced incidents related to AI agents, where access credentials have been compromised. This raises a crucial question: how can we ensure security while leveraging these advanced technologies?<\/li>
  • A series of incidents has been documented:<\/li>
  • Type of incident<\/li><\/ul>\n\n

    Percentage of companies affected<\/h3>\n\n

    Internal information leaks<\/p>\n\n

    38%<\/p>\n\n

    \n\n\n\n\n\n
    External attacks on AI systems<\/th>\n29%<\/th>\n<\/tr>\n<\/thead>\n
    Governance often insufficient<\/td>\nDespite recognizing the importance of strong governance, only 44% of companies have implemented dedicated policies for AI agent management. This leads to a situation where the majority of access information is delegated to the IT department, creating a lack of collaboration between technical teams and legal or compliance departments. This situation is worrying because it can lead to insufficient oversight of agents and an erosion of customer trust. The Importance of an Identity-Centric Approach<\/td>\n<\/tr>\n
    To counter these threats, experts like Chandra Nanasambandham of SailPoint advocate for rigorous identity management. An approach based on transparent governance of AI agents is essential to prevent abuse and ensure security.<\/td>\nThe Principle of Least Privilege<\/td>\n<\/tr>\n<\/tbody>\n<\/table><\/figure>\n\n

    An effective strategy is to apply the principle of least privilege, ensuring that AI agents only access the information necessary to perform their functions. This principle is often overlooked, especially given the rapid pace of access implementation.<\/h3>\n\n

    Here are some steps for effective identity management:<\/p>\n\n

    Continuously assess agent access levels.<\/h2>\n\n

    Implement strict access control policies.<\/p>\n\n

    Train teams on AI agent security.<\/h3>\n\n

    Cohesion between departments for better security<\/p>\n\n

    Collaboration between IT, legal, and compliance departments is essential. It’s about establishing a network where information flows freely, enabling rapid response to potential threats. A framework for consultation between these departments is essential.<\/p>\n\n

    • A look back at case studies and best practices<\/li>
    • To offer concrete solutions, let’s examine some case studies of companies that have effectively managed the risks associated with AI agents while ensuring robust cybersecurity.<\/li>
    • Case study: Sopra Steria and cybersecurity<\/li><\/ul>\n\n

      Sopra Steria, a key player in the world of digital services, recently strengthened its cybersecurity strategy by integrating advanced AI agents. Focusing on strong governance, the company implemented a system to monitor the activities of these agents in real time.<\/h3>\n\n

      Best practices include:<\/p>\n\n

      Regular auditing of AI systems to detect vulnerabilities.<\/h2>\n\n

      Implementing attack simulations to test resilience. Proactive approach to educating employees on cybersecurity issues.<\/p>\n\n

      Collaboration with security partners such as Kaspersky<\/h3>\n\n

      Other companies, such as Gemalto, Atos, and Kaspersky, are collaborating to develop secure solutions incorporating the latest advances in AI. Their strategy is based on anticipating and managing risks through advanced encryption and authentication technologies.<\/strong>The benefits of such collaborations include:<\/p>\n\n

      Better real-time threat detection.<\/p>\n\n

      • Accelerated responses to security incidents.<\/li>
      • Improved attack prevention capabilities.<\/li>
      • Preparing for future cybersecurity challenges with AI<\/li><\/ul>\n\n

        The challenges posed by the rise of AI agents in the enterprise are clear. According to the report, nearly half of companies cannot track or monitor how these agents use their data. Given this reality, it is essential to prepare for the future.<\/h3>\n\n

        Strategies for Proactive Cybersecurity Industry leaders such as Sentryo and Stormshield advocate a proactive approach, including the integration of predictive analytics tools to anticipate threats. Companies should follow several guidelines:<\/strong>Invest in advanced intelligence tools for activity monitoring. Adopt adaptive technologies to respond instantly to anomalies.<\/strong> Collaborate with experts such as Wavestone to develop tailored cybersecurity strategies. The Role of Education and Awareness<\/strong>Finally, educating employees about the risks associated with AI agents and security best practices is crucial. Everyone in an organization must understand their role in data security. This includes:<\/p>\n\n

        Regular training sessions on AI tools and their impacts.<\/p>\n\n

        • Provision of educational resources on cybersecurity.<\/li>
        • Awareness campaigns on the most common attack patterns.<\/li>
        • <\/li><\/ul>\n\n

          <\/h2>\n\n

          <\/p>\n\n

          <\/h3>\n\n

          <\/strong> <\/strong><\/p>\n\n

          • <\/li>
          • <\/li>
          • <\/strong> <\/li><\/ul>\n\n

            <\/h3>\n\n

            <\/p>\n\n

            • <\/li>
            • <\/li>
            • <\/li><\/ul>\n\n\n","protected":false},"excerpt":{"rendered":"

              As we approach 2025, businesses have never faced more cybersecurity challenges, particularly with the emergence of artificial intelligence (AI) agents. These promising tools, which simplify many processes, can also become vectors for unsuspected threats. The SailPoint report, released in July 2025, highlights a worrying reality: the growing adoption of AI agents is accompanied by new […]<\/p>\n","protected":false},"author":1,"featured_media":46595,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[41],"tags":[2395,16473,1413,4857,80459],"class_list":["post-46605","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-non-classe-en","tag-ai-agent-en","tag-companies-en","tag-computer-security-en","tag-cybersecurity-en","tag-discreet-threat-en"],"_links":{"self":[{"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/posts\/46605","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/comments?post=46605"}],"version-history":[{"count":1,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/posts\/46605\/revisions"}],"predecessor-version":[{"id":46606,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/posts\/46605\/revisions\/46606"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/media\/46595"}],"wp:attachment":[{"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/media?parent=46605"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/categories?post=46605"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.mon-agent-ia.fr\/blog\/wp-json\/wp\/v2\/tags?post=46605"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}