Cybersecurity professionals these days have the daily task of determining how best to leverage AI to enhance cyber resilience. AI-powered tools are increasingly a top priority. In fact, 64% of cybersecurity experts are looking to integrate these tools into their existing systems to bolster defenses against AI-driven threats.
Channel partners may be unsure of how AI-powered tools fit within their tech offerings and cybersecurity strategies in the long term. Sure, it’s tempting to push AI tools as a remedy to growing AI threats. However, keep in mind that human error is the leading cause of cyberattacks. In fact, social engineering attacks prey upon this vulnerability. AI further enhances these tactics. Fighting AI with AI will not limit human error.
The Rising Threat of AI-powered Social Engineering
Personalizing phishing campaigns and scanning the web for exploitable information is one way AI makes them more believable. The technology can even streamline the process with automation to reduce the cost for attackers while increasing the reach and impact of campaigns.
One way is deepfakes, which can convince employees to hand over credentials or money by impersonating their colleagues. For example, an employee in Hong Kong received an email from their company’s CFO about a confidential transaction. The employee attended a virtual meeting, where they saw the CFO as well as other colleagues. But after approving the transaction, they discovered it was an AI-driven deep fake. This scenario led to a $25 million loss, underscoring how AI-enabled social engineering can be a high-impact, easy-to-deploy threat.
In the fight against AI-powered attacks, the channel cannot solely rely on AI-powered tools. It is imperative that MSPs guide organizations in limiting the vulnerabilities posed by human errors to build robust cyber resilience.
The Human Factor: An Ongoing Challenge
The fundamental weakness in any system is still the human element. Identity attacks are on the rise, exploiting credentials like passwords, API keys, or browser cookies. Microsoft’s recent report identified more than 600 million identity attacks in its 2024 fiscal year alone.
As organizations scale, the likelihood of human error increases — and attackers know this all too well. So, MSPs must work with customers to eliminate human error, as these are prime targets for social engineering attacks.
Moving Beyond Identity Silos
The issue of passwords becomes even more critical as IT infrastructure grows more complex. Credentials are scattered across myriad systems, including Kubernetes clusters, cloud APIs, IoT devices, specialized dashboards, and databases creating “identity silos.” When organizations experiment with AI, they often introduce AI agents that require access to these systems. These AI agents themselves become security silos that act as backdoors, vulnerable to exploitation if not managed properly.
Ev Kontsevoy
For example, an attacker could gain access by targeting the credentials of an AI agent to breach employees. They could access sensitive data the AI uses for training, creating a significant data breach. Without effective identity access frameworks, finding the source of such an attack can become a needle-in-a-haystack situation.
The Case for Consolidating AI with the Rest of Your Identity
Eliminating static and standing credentials go hand in hand with unified identity governance. To fight AI threats, channel partners must shine a light on the solutions that bolster security and safeguard the most frequent target of cyberattacks: compromised identities.
Mitigating compromised credentials risks requires organizations to treat AI agents with the same security protocols as other resources in their environment. This means consolidating identities across all systems — laptops, servers, databases, microservices — into a single, centralized inventory that provides a comprehensive view of identity and access relationships.
However, consolidation alone isn’t sufficient. Organizations must adopt dynamic, cryptographic methods of authentication. Cryptographic identity, for instance, uses key pairs for secure, verifiable digital signatures. This can reduce the vulnerabilities associated with passwords. It employs a decentralized model that minimizes points of failure and can integrate with technologies like blockchain for added transparency. With cryptographic identity, companies can ensure that access is granted on a task-by-task basis for a limited duration, significantly reducing the blast radius and the risk of unauthorized access.
A good model for security already exists in the iPhone. It combines a biometric marker (face recognition), a personal identification number (PIN), and the device’s trusted platform module (TPM) chip. Applying a similar three-point authentication system for employees — using device identity, biometrics, and a PIN — can greatly enhance security.
Balancing Security with AI-powered Efficiency
Consolidating identities isn’t just safer; it can also lead to productivity gains. Streamlining identity governance can reduce the friction between maintaining robust security and adopting innovative technologies, making it a valuable selling point for MSPs advocating for a unified strategy.
As organizations expand their tech environments, consolidating identities in a single source of truth and implementing cryptographic identity are practical steps to ensure security and efficiency.
Ev Kontsevoy is co-founder and CEO of Teleport. A serial entrepreneur with a background in engineering, he previously co-founded and led Mailgun, later acquired by Rackspace.
Featured image: DALL-E
