This has been one of the busier weeks for threats affecting UK organisations. Several critical issues are being actively exploited, and at least one has no fix available yet.
The most pressing items involve Cisco and SolarWinds products widely used in NHS and enterprise environments, including a flaw in Cisco's phone systems that already has an NHS cyber alert and a public exploit in circulation. A supply chain worm called Miasma has also escalated significantly, now hitting 73 Microsoft GitHub repositories and spreading through open-source tools used by development teams. Elsewhere, an AI agent uncovered 21 long-hidden vulnerabilities in FFmpeg, Google pushed a record Chrome patch release, and a third-party breach at Oxford University exposed student and alumni data.
Full details and recommended actions for each are below...
Cisco has disclosed a critical vulnerability in its Unified Communications Manager (Unified CM) and Unified CM Session Management Edition products. The flaw is tracked as CVE-2026-20230 and is a server-side request forgery (SSRF) vulnerability. In plain terms, that means an attacker can send specially crafted requests to an affected system over the network, without logging in, and trick the server into writing files to its own operating system. Those files can then be used to escalate privileges all the way to root, meaning the attacker gains full, unrestricted control of the server.
A proof-of-concept exploit for this vulnerability is already publicly available, meaning anyone can download a ready-made tool to run the attack. NHS England's National Cyber Security Operations Centre (CSOC) has assessed exploitation as likely. The WebDialer service must be enabled for the server to be vulnerable. This is disabled by default, but many enterprise deployments enable it.
Cisco Unified Communications Manager is the phone system backbone used across many NHS trusts, NHS supplier organisations, and UK enterprise environments. It handles calls, voicemail, and unified messaging for thousands of staff. A successful exploit gives an attacker root-level control of the communications server, which can be used to intercept calls, steal credentials, or move laterally into connected systems. For organisations with DSPT obligations, a compromise of a system handling staff or patient communications is a serious incident that may require reporting. The NHS has specifically flagged this as a priority for NHS-connected organisations.
Recommendations
A self-replicating software supply chain worm called Miasma has now compromised 73 GitHub repositories belonging to Microsoft, including repositories across Microsoft's Azure, Azure-Samples, Microsoft, and MicrosoftDocs organisations. GitHub has disabled access to the affected repositories. The repositories include key Azure developer tools such as azure-functions-host, durabletask, and various cloud connector components.
Miasma is a variant of the Shai-Hulud worm that began spreading through the npm open-source package registry in May 2026. It spreads by stealing the credentials of legitimate software publishers and then publishing malicious versions of packages under those trusted accounts. Because the malicious packages are signed with valid keys and published by authenticated accounts, they look completely legitimate to package registries and security tools. In some cases, the worm has been observed planting a malicious payload that executes automatically when a developer opens an infected project in an AI coding assistant such as Claude Code, Gemini CLI, Cursor, or VS Code.
The attack has also hit Red Hat npm packages and a range of other widely used open-source libraries with hundreds of thousands of monthly downloads.
NHS suppliers, digital health companies, and healthtechs that build software using open-source components from npm, PyPI, or GitHub are all potentially exposed. If a development team installs or uses a Miasma-infected package, the worm can steal credentials from the developer's machine and use them to compromise further repositories, creating a chain reaction. This could result in malicious code being inserted into software products that are later deployed in NHS or clinical environments, a classic supply chain attack. The risk does not require the developer to do anything wrong; installing a seemingly legitimate package is enough. Organisations that rely on third-party developers or open-source libraries in their products should be asking their suppliers about this now.
Recommendations
CISA has added a high-severity vulnerability in SolarWinds Serv-U to its Known Exploited Vulnerabilities (KEV) catalogue, meaning it has confirmed evidence of active exploitation in the wild. The flaw is tracked as CVE-2026-28318 and has a severity score of 7.5 out of 10.
Serv-U is a multi-protocol file server product used to send and receive files securely, often used for managed file transfer (MFT) in enterprise and healthcare settings. The vulnerability is a denial-of-service flaw; an attacker can send a specially crafted HTTP POST request to the Serv-U service without needing a username or password. This causes the service to crash. When Serv-U crashes, any organisation relying on it for secure file transfer loses access to that function immediately.
A patch is available in Serv-U version 15.5.4 HF1. CISA has ordered US federal agencies to apply the fix by 19 June 2026. SolarWinds products have a history of serious vulnerability exploitation, including by the Cl0p ransomware gang, which previously targeted earlier Serv-U flaws as an initial access route into organisations.
Serv-U is widely used across UK enterprise environments, including NHS suppliers, for the secure transfer of documents, reports, and data files. Managed file transfer services are critical in healthcare settings where lab results, imaging reports, referral letters, and contract documents are shared between organisations. A successful denial-of-service attack against Serv-U would stop all file transfers immediately, potentially disrupting clinical workflows. Given the history of ransomware groups targeting Serv-U, there is a realistic risk that exploitation of this flaw is a precursor to more serious intrusions. For DSPT-registered organisations, an outage affecting the transfer of patient data may need to be logged even if no data is actually lost.
Recommendations
Cisco has issued a warning that a high-severity vulnerability in Catalyst SD-WAN Manager is being actively exploited in the wild. The flaw is tracked as CVE-2026-20245 and has a severity score of 7.8 out of 10. At the time of publication, there is no patch available.
SD-WAN Manager (formerly called vManage) is the central management platform for Cisco's SD-WAN (software-defined wide-area networking) products. The vulnerability is in the command-line interface of the product. An attacker who already has local access, for example, an insider threat, or an attacker who has gained initial access through another route, can upload a specially crafted file that the system processes without properly checking its contents. This allows the attacker to run commands as the root user, giving them complete control of the SD-WAN management system. The flaw affects on-premises deployments, Cisco SD-WAN Cloud-Pro, Cisco-managed cloud deployments, and FedRAMP-authorised environments.
Cisco has confirmed active exploitation and said there are no workarounds that fully remove the risk.
SD-WAN technology is used by enterprise-scale NHS trusts, NHS suppliers, and digital health organisations to manage the network connections between sites, data centres, and cloud environments. The SD-WAN Manager is the single control plane for those networks, meaning a compromise of the management platform could allow an attacker to reroute traffic, disable connections, or pivot into any part of the network it manages. The fact that this is being actively exploited with no patch available is a serious concern. Any organisation using Cisco SD-WAN should assume this vulnerability will be targeted and take immediate steps to reduce the risk of exploitation. For NHS-connected organisations, a network management system compromise could have a direct impact on clinical connectivity and service availability.
Recommendations
A security startup called depthfirst used an autonomous AI agent to scan FFmpeg, the open-source media library embedded in a huge proportion of video-processing software worldwide and found 21 previously unknown vulnerabilities. Each one was confirmed with a working proof-of-concept. Several of these bugs had been sitting undetected in the code for between 15 and 23 years. Most are memory corruption flaws in the parts of FFmpeg that handle media file formats, video decoders, and transport streams. Nine CVEs have been assigned so far (CVE-2026-39210 through CVE-2026-39218), with more to follow. The entire scan cost approximately $1,000 to run.
In the same week, Google released Chrome 149 with patches for a record 429 security bugs, more than any previous Chrome release. Over 100 are rated critical or high severity. The most severe, CVE-2026-10881 (CVSS 9.6), is a memory flaw in Chrome's ANGLE graphics engine that allows a malicious web page to escape the browser's sandbox and run code directly on the user's device. Google paid $97,000 to the researcher who found it. Chrome's auto-update mechanism should apply the patch automatically, but users should confirm they are running version 149.0.7827.53 or later.
FFmpeg is embedded in a very wide range of products, including media players, video conferencing tools, streaming servers, diagnostic imaging software, and countless web applications. Many of these embedded copies are not updated when the standalone FFmpeg project ships a patch. If your organisation or any of your suppliers uses software that processes video, audio, or media files, there is a good chance FFmpeg is inside it somewhere. Some of the discovered bugs are exploitable via malformed media files, meaning an attacker could potentially trigger them by sending a crafted video to an affected system. In a healthcare context, this matters anywhere medical imaging, video consultation, or media processing is in use. The Chrome vulnerability is a reminder that browsers used by clinical and administrative staff remain one of the most common entry points for attackers.
Recommendations
Oxford University has disclosed a data breach affecting its CareerConnect platform, which is operated by third-party provider Group GTI and used by students, alumni, research staff, and external recruiters to find work opportunities. The breach took place on 28 May 2026, when attackers exploited a security vulnerability in the platform. The vulnerability has since been fixed.
The data exposed includes full names and email addresses for all affected users. Those who did not use single sign-on (SSO), meaning they had a separate password for the platform, also had their encrypted passwords leaked. Oxford has confirmed that course information, uploaded files, appointment information, and financial data were not involved. This is the second time in as many months that Oxford has been affected by a breach at an external platform provider.
This incident is a direct example of third-party platform risk — a risk that is highly relevant to NHS-connected organisations, universities with NHS research partnerships, and any organisation that provides career services, recruitment support, or staff development platforms through an external supplier. The breached data — names, email addresses, and encrypted passwords — is exactly what attackers use to build phishing campaigns and attempt credential stuffing attacks against other platforms. Staff and alumni who reuse passwords across services are at particular risk. For organisations subject to DSPT obligations, a breach involving personal data held by a third-party processor is still a reportable event, even if the attack happened in the supplier's environment rather than your own.
Recommendations
Want help staying ahead of threats like these? Contact Periculo about our Threat Intelligence services and find out how we support UK digital health organisations, healthtechs, and NHS suppliers with practical, hands-on cybersecurity assurance.