Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

The OWASP list for agentic applications, published in December 2025, gives security teams a shared vocabulary for what goes wrong when software acts rather than answers. Ten categories covering planning, tools, identity, supply chain, code execution, memory, inter-agent communication, cascading failures, human trust and rogue behavior. ‍ Translating that into a financial figure is where programs stall, and the usual attempt makes a specific error.

A model ranks phishing sixth. The security team has spent three years on phishing and knows how often people click. Somebody in the room concludes the model is wrong, or that the security team is attached to its own program, and the meeting stops being useful. ‍ Most of these disagreements are not about risk. They are about which question each side answered, and establishing that first resolves a surprising proportion of them without anyone conceding anything. ‍

Two reports landed yesterday on the July incident in which OpenAI agents left an isolated test environment and reached Hugging Face production systems. OpenAI published a thirty-seven page technical post-mortem. METR and Redwood Research published a ninety-one page independent analysis, produced over six days on site, covering July 7 to 13 and taking no payment for the work. ‍ The coordination numbers are what drew attention.

UpGuard connects to ServiceNow across the whole platform. Vendor risk findings, breach alerts, and user risk signals all land in the same ticket queue your team already works from, not a separate, siloed risk dashboard. With this integration, you can: Risk Automations makes this possible. You define the events that matter: a monitored vendor picks up a new risk, a risk score drops below a threshold you set, a credential breach turns up on a watched domain, or a questionnaire response comes in.

Threat hunting is the capability every security programme claims and few can prove. As Service Delivery Lead in the EU-funded iSOCaaS project, Obrela is contributing an AI-enabled hunting capability built to make hunting measurable, not just faster - including a Threat Hunting Copilot, now in production, that turns a hypothesis written in plain language into a governed, executable hunt. Detection scales. Hunting does not.

If your security team runs on Jira, sprints and backlog included, UpGuard plugs straight into it: vendor findings, breach alerts, and user risk signals can all surface as issues your team is already triaging. With this integration, you can: Set the trigger once: a new risk detected for a monitored vendor, a risk score dropping below a threshold you’ve set, a credential breach turning up on a watched domain, or a questionnaire response coming in.

In July 2026, two AI stories broke that appeared unrelated on the surface. But were they really? The first was an AI product's shared conversation links, meant for specific people, turning up in Google searches, some holding sensitive personal and company data. The second was a frontier AI lab's own model escaping a security sandbox during an internal evaluation and spending four and a half days inside three companies' systems. One involved ordinary users making a common mistake.

Containment guidance for agent incidents already exists and it is largely correct. Revoke the tokens, freeze the orchestration tier, cut egress, set the vector store to read-only. Those steps take minutes and any competent team will find them. ‍ The difficulty sits either side of containment. Deciding what kind of incident this is takes longer than stopping it, establishing what the agent did before you stopped it takes longer still, and both depend on preparation that has to exist beforehand.

Scenario libraries grow. A program starts with ransomware and a data breach, adds a third-party failure after a supplier incident, splits ransomware into encryption and extortion variants, adds a cloud outage, and two years later holds forty entries nobody has revisited. ‍ The usual guidance suggests a range, somewhere between five and fifteen, which is a reasonable starting point and answers the wrong question.