The growing presence of artificial intelligence (AI) and machine learning (ML) in various industries has raised concerns about security and identity governance. Recent breaches, such as the ShinyHunters' exploitation of an Oracle zero-day vulnerability in higher education institutions, have highlighted the need for better control over AI agents and their interactions.
What Happened
In June 2025, Simon Willison, the engineer who coined the term "prompt injection," published a warning that circulated widely through the security community. He called it the lethal trifecta — three capabilities that, when combined in a single AI agent, create a near-guaranteed path to exploitation through indirect prompt injection: access to private data; exposure to untrusted content; the ability to communicate externally.
The same class of attack has been seen in various production exploits, including Microsoft 365 Copilot, GitHub's MCP server, GitLab Duo, Slack AI, Google Bard, and Amazon Q. These attacks have raised concerns about the security of AI systems and the need for better identity governance.
Why It Matters
The security industry has had plenty of warnings about the risks associated with AI and machine identities, but it has not acted on them. The SolarWinds story is a prime example, where attackers slipped in, found machine identities with significant access, and used them without being detected for months. This has led to the concept of "ghost identities" — non-human identities that outnumber human ones in most large organizations and can accumulate privilege and wait for exploitation.
What Experts Say
"The real control point turned out to be somewhere else entirely." — Expert on sovereign cloud and AI risk
Experts argue that sovereign cloud alone is not enough to mitigate AI risks. Instead, identity governance is key to controlling AI workloads and ensuring security. European enterprises have already discovered this under real regulatory pressure, with many migrating workloads, renegotiating contracts, and writing sovereign cloud into board-level risk frameworks.
Key Numbers
- 42% of organizations have experienced a breach due to compromised AI agents
What Comes Next
As AI adoption continues to grow, it is essential to prioritize identity governance and security. This includes implementing robust controls over AI agents, monitoring their interactions, and ensuring that they are not compromised. The recent breaches serve as a wake-up call for organizations to take proactive measures to protect themselves against AI-related risks.