Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Bring AI out of the shadows with visibility and control that help your organization adopt AI securely. When we introduced Sophos AI Defense in July, we focused on a challenge facing organizations everywhere: AI adoption is accelerating, often faster than security and IT teams can see or control it. New AI tools, embedded capabilities, and shadow AI are creating blind spots around where AI is being used and what information it can access.

EDR Killers let attackers blind endpoint protection before the real attack begins, and Sophos Endpoint stops them with several layers that are on by default. Most people picture a cyberattack as something loud: files encrypting, screens locking, a ransom note. But the attacks that do real damage usually start quietly. Before an attacker steals data or deploys ransomware, they try to make sure nobody is watching. That means dealing with the endpoint protection on the device first.

This article was first published on LinkedIn. In Minority Report, the Department of Precrime promised certainty. Three precognitives saw the future, the system picked the consensus, and police arrested people for murders they hadn't yet committed. The whole machine depended on one buried secret: sometimes the precogs disagreed, and the dissenting vision, the minority report, was quietly deleted. The system held together as long as the precogs agreed, but broke when it removed its own doubt.

I’ve watched AI-powered cyberattacks inch closer to autonomy throughout 2026, and I can point to dozens of real-world incidents as proof. AI-enabled cyberattacks surged by 56% over the last year, meaning 1 in 4 (25%) malicious breaches now involve artificial intelligence. The important one: on May 10, 2026, researchers documented the first fully autonomous LLM-driven post-exploitation attack in the wild.

Six AI agent security best practices for teams running coding agents and MCP servers: limit what a compromised agent can reach and catch credential misuse.

CVE-2026-89110 is a broken access control vulnerability in SAP NetWeaver Visual Composer, classified as CWE-749, Exposed Dangerous Method or Function. Certain portal functionality does not perform sufficient authorization checks, and a low-privileged attacker can use it to invoke internal operations on a preconfigured remote system. The vulnerability carries a CVSS v3.1 base score of 9.9 (Critical), with the vector AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H.

We talk about the cloud as though data somehow disappears into an abstract environment, but the cloud is still physical. Somewhere there is a building filled with power systems, cooling, fiber, servers, cameras, access control and people. As data center operators expand across multiple locations, cloud-managed physical security gives them a way to manage access, video and identity consistently across the enterprise while still maintaining local resilience.

Nearly half of the organizations we surveyed — 49.8% — confirmed or suspected that an AI agent in their environment took an action in the past year that nobody intended, expected, or authorized. That is not a prediction. It is what security and technology professionals told us when we asked them directly, across a survey of over 380 organizations, about their experience over the last 12 months. Agentic AI has moved fast. The data shows it has moved faster than most organizations can govern.

Open banking drives financial innovation but introduces new cybersecurity risks through APIs, exposed credentials and third-party integrations. Discover how cyber threat intelligence helps financial institutions identify emerging threats, improve visibility and strengthen security.

Monitoring has to happen on the device, because three categories of AI use never reach the network in a form anything can inspect. Local models run without leaving the machine. Desktop applications that pin their certificates sit on an inspection bypass list by design. A laptop on a home connection routes through nothing the organization operates at all. ‍ The uncomfortable part is that the first two are deliberate. Somebody approved the bypass, and the approval was correct on its own terms. ‍