Loading story…
Gartner analyst Wam Voster warned that agentic AI in operational technology (OT) environments could create new cyber-physical and safety risks as autonomous systems begin making production decisions in industrial settings. He said factories may increasingly rely on AI agents to adjust variables such as temperature, timing, and equipment settings based on external and process data, raising the possibility that configuration errors or unsafe automated decisions could move beyond downtime and production loss into physical harm.
Voster highlighted that a misconfigured AI agent could make dangerously incorrect adjustments—for example, changing a process by 200 degrees instead of 2 degrees—with potential consequences including equipment destruction, explosions, or injury and death. He also emphasized that human error remains a major cause of cyber-physical incidents and discussed the limits of applying full zero trust concepts in industrial environments. A separate ZDNET workplace advice article about employee anxiety over AI agents is not about OT security or this specific risk discussion and should be excluded.
Track how attackers are adapting to this technology.
1 event from the most recent confirmed update back to the earliest known activity.
Gartner vice president analyst Wam Voster warned that agentic AI in industrial and operational technology environments could autonomously change process variables such as temperature, timing, and equipment settings, creating serious safety and security risks if misconfigured. He highlighted governance, asset visibility, human error, and the practical limits of zero trust in industrial settings as key concerns.
Follow how adversaries are adapting to this technology, and where it touches your stack today.
2 references tracked. Mallory keeps watching after this page renders.
Map indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
Gartner warned that **misconfigured AI embedded in cyber-physical systems** could trigger a shutdown of **national critical infrastructure** in a G20 country by 2028, driven by AI failures rather than malicious hacking. The scenario focuses on AI-enabled operational environments where automated decision-making and control can propagate errors rapidly from digital systems into physical processes. The risk is framed around **engineering and configuration mistakes**—such as flawed update scripts, incorrect parameters (e.g., a misplaced decimal), or misinterpreted sensor data—causing unsafe actions, equipment damage, cascading outages, and supply-chain destabilization. Gartner highlighted that modern critical environments (including **operational technology/ICS**, industrial automation, IIoT, robots, drones, and AI-assisted power grid balancing) are increasingly reliant on AI, making resilience, validation, and change-control for AI-driven control logic a critical safety and security concern.
Jun 29, 2026
Security leaders are warning that **AI agents are increasingly operating as “digital employees”** inside enterprise workflows—triaging alerts, coordinating investigations, and moving work across security tools—often with **broad permissions and limited governance**. The core risk highlighted is that organizations are deploying high-authority agents like plug-ins (reused service accounts, overbroad roles, weak oversight), creating fast-acting operators that can be manipulated and that lack the contextual judgment and policy awareness expected of human staff. Related commentary also raises concerns about **AI-to-AI communication** and “non-human-readable” behaviors that could reduce auditability and complicate investigations and control enforcement. In parallel, public examples show how quickly AI can accelerate **vulnerability discovery**: Microsoft Azure CTO Mark Russinovich reported using *Claude Opus 4.6* to decompile decades-old Apple II 6502 machine code and identify multiple issues, underscoring that similar techniques could be applied to **embedded/legacy firmware at scale**. Anthropic has also cautioned that advanced models can find high-severity flaws even in heavily tested codebases, reinforcing the likelihood that both defenders and attackers will leverage AI for faster bug-finding. Separate enterprise IT coverage notes that organizations are **reallocating budgets toward AI** by consolidating tools and renegotiating contracts, which can indirectly increase security exposure if cost-cutting reduces overlapping controls or if AI adoption outpaces governance and identity/access management maturity.
Jun 29, 2026
Organizations are increasingly blending **agentic AI** with deterministic software, but the primary risk is operational: probabilistic systems can produce confident, plausible outputs that are not reliably predictable, auditable, or repeatable. Guidance aimed at CIOs emphasizes using traditional code for **data validation, transaction processing, compliance logic, and structured output generation**, while reserving agents for ambiguous tasks like interpretation, summarization, and decision support; it also notes that teams should tune agent behavior and output quality before adding guardrails, because weak prompting or ill-suited models can shift safety boundaries. Gen’s research frames this as **“Artificial Mindless Intelligence (AMI)”**—AI that generates mind-like output without grounded understanding, stable intent, or accountability—and warns that harm does not require a model to “go rogue.” The key failure mode is granting AI systems **operational authority** (e.g., access to inboxes, files, calendars, chat apps, admin consoles, or payment workflows), where errors, manipulation, or ungrounded actions can directly trigger business impact; the piece argues that focusing on AGI narratives leads either to security theater or complacency, while the near-term risk is permissioned automation acting beyond appropriate authorization boundaries.
Jun 29, 2026