Nvidia's Neocloud Faces Power Delay; GitLab Fixes Critical Vulnerability

Nvidia-backed Neocloud Faces Power Supply Delays

Nvidia-backed Neocloud's AI supercomputer, developed by Nscale, is facing significant delays due to power supply issues in the UK. Initially slated for a 2027 launch, the Loughton data center in Essex requires a massive 90MW of power, equivalent to the energy needs of 315,000 homes. However, the UK's National Grid is currently overwhelmed, with a backlog of over 300 data centers needing power, which collectively demand 73GW, surpassing the UK's capacity of 45GW by 62%.

The Neocloud project is a GPU-based AI cloud data center, representing Nscale's inaugural venture. The startup is preparing for a £26 billion public listing and has already secured £2.55 billion in pre-IPO funding. The data center's operation hinges on securing the necessary power supply, a challenge echoed across the UK, where numerous data center projects are stalled due to similar issues.

The implications of these delays are significant for the UK's AI ambitions. As the demand for AI infrastructure grows, the UK's current energy strategy, which has shifted focus towards renewable resources, is proving insufficient. This has prompted some data centers to consider building their own gas-fired power stations, highlighting a potential shift back towards more traditional energy sources.

From a broader perspective, the power supply challenges facing Neocloud underscore the critical need for scalable and sustainable energy solutions in supporting the next wave of AI infrastructure. The situation also raises questions about the UK's readiness to support large-scale AI projects and may influence future investment decisions in the region.

Critical Vulnerability in GitLab's AI Gateway

GitLab has patched a critical vulnerability in its AI Gateway, identified as CVE-2026-90970, which had a CVSS score of 9.9. This flaw allowed authenticated users with access to the Duo Agent Platform to execute arbitrary commands on the gateway. The vulnerability was disclosed on October 2, and GitLab released fixes in AI Gateway versions 19.2.4, 19.3.2, and 19.4.1.

The vulnerability stemmed from the AI Gateway's handling of custom flow prompt templates. Users with access to the Duo Agent Platform could craft a flow configuration that escaped the prompt template sandbox, leading to command execution on the AI Gateway host. GitLab credited the HackerOne researcher 'invisiblemeerkat' for responsibly reporting the issue.

This vulnerability highlights the complexities of securing AI infrastructure, particularly as organizations increasingly deploy AI Gateways within their own environments. The incident underscores the importance of robust security practices in AI deployments, as the gateway has access to sensitive authentication materials and interacts with AI models.

For the industry, this incident serves as a reminder of the critical importance of securing AI infrastructure against potential vulnerabilities. As AI systems become more integral to business operations, ensuring their security will be paramount to maintaining trust and functionality.

ShinyHunters Hacker Detained in Jordan

A member of the ShinyHunters hacking group, known online as 'Rey', has been detained in Jordan and is cooperating with the FBI. Identified as Saif al-Din Khader, Rey was taken into custody earlier this week and is assisting law enforcement in locating other members of the extortion group.

ShinyHunters has been involved in several high-profile cyberattacks, including a breach of the FBI's systems using an alleged Oracle PeopleSoft zero-day vulnerability. The group claimed to have stolen between 2TB and 3TB of data, including sensitive information from FBI employees and applicants.

The detention of Rey marks a significant development in the ongoing efforts to dismantle ShinyHunters. The group's operations have been disrupted, with their data leak site going offline and communications ceasing following Rey's arrest. However, a new data leak site has since appeared, indicating that other members may continue their activities.

This development highlights the ongoing challenges in combating cybercrime, particularly against organized groups like ShinyHunters. The cooperation between international law enforcement agencies and the use of advanced cyber forensics are crucial in addressing these threats and bringing perpetrators to justice.

Where Robles.AI Fits

Robles.AI specializes in providing scalable and secure AI solutions, which are directly relevant to the challenges highlighted in these articles. For instance, our cloud-based AI infrastructure solutions can help address the power supply issues faced by large-scale AI data centers like Neocloud. By optimizing energy efficiency and integrating renewable energy sources, we can support sustainable AI growth.

In terms of cybersecurity, Robles.AI's advanced threat detection and response capabilities are designed to protect AI gateways and infrastructure from vulnerabilities similar to those identified in GitLab's AI Gateway. Our solutions ensure that AI deployments remain secure and resilient against potential threats.

Furthermore, Robles.AI's expertise in AI-driven cybersecurity can assist in tracking and mitigating the activities of cybercriminal groups like ShinyHunters. By leveraging AI to analyze and predict cyber threats, we can enhance the security posture of organizations and support law enforcement efforts in combating cybercrime.