Curated intelligence for federal IT decision makers.

CybersecurityThe Hacker News · 2026-08-28

China-Made ZBT Routers Contain Implants Allowing Root Access to Unauthenticated Attackers

Security researchers have uncovered two hidden implants in firmware for routers manufactured by Shenzhen Zhibotong Electronics (ZBT), enabling unauthenticated remote attackers to execute commands with root-level access. These implants, named SPEAKINGSTONE and DARKLANTERN and tracked as CVE-2026-74232 and CVE-2026-74233, were discovered by VulnCheck. Both vulnerabilities scored 9.3 on the CVSS 4.0 scale and 9.8 on CVSS 3.1, indicating their critical severity.

SPEAKINGSTONE operates as the service yunmgrd and sends UDP beacons to a hardcoded command-and-control (C2) server, allowing attackers to execute arbitrary commands, exfiltrate credentials, and hijack DNS. DARKLANTERN, running as infosrvd, opens UDP port 9992 to inbound connections without effective authentication. VulnCheck identified 203 internet-facing DARKLANTERN instances across 22 countries between August 18-21, 2026.

The implants were found on an $88 Deep Orange router purchased from a U.S. supplier, which uses ZBT firmware built in 2019. Multiple ZBT router models are affected, including WE826-T2, WE2426-C, and L3_V2_8. VulnCheck has not yet identified fixed firmware versions, leaving many devices vulnerable. The company confirmed the implants ship with ZBT firmware, though MOFI Network’s custom firmware was found to be unaffected.

VulnCheck recommends blocking traffic to UDP ports 9992 and 10000, treating router LANs as untrusted, and monitoring for Indicators of Compromise (IoCs). ZBT has not publicly addressed these vulnerabilities, though its firmware download pages remain active as of August 28, 2026. The discovery highlights supply chain risks in network hardware.

• ZBT routers contain two implants allowing unauthenticated root access.
• SPEAKINGSTONE and DARKLANTERN vulnerabilities scored 9.3 and 9.8 on CVSS scales.

CybersecurityThe Hacker News · 2026-08-28

Identity Fabric Architecture Gains Importance for Enterprise Security in 2026

Enterprise identity security is evolving beyond static configurations toward runtime visibility as hybrid and multi-cloud environments create fragmented access paths. An Identity Fabric architecture bridges the gap between access policy intent and actual execution by connecting identity providers, governance systems, and infrastructure into an observable layer.

Traditional identity management operates across design-time (provisioning, policy definition) and runtime (authentication, authorization) dimensions. The unobserved space between these dimensions—termed 'identity dark matter'—allows risks like credential sprawl and privilege escalation to accumulate. Modern environments exacerbate this challenge with proliferating SaaS integrations, API-to-API authentication, and automated workloads that bypass traditional governance controls.

Non-human identities (service accounts, bots, cloud workloads) now represent a majority of enterprise credentials yet receive minimal governance attention. Overprivileged, dormant, or unowned machine identities create attack surfaces, especially control-plane credentials that can modify infrastructure security settings. Effective management requires assigning ownership, defining purpose-limited permissions, enforcing expiration, and monitoring behavioral deviations.

Identity Fabrics support Zero Trust by providing continuous access evaluation against actual usage patterns rather than periodic reviews. Unified visibility across hybrid environments helps detect cloud lateral movement through IAM trust relationships, while behavioral baselines accelerate incident response by correlating activity across systems. Emerging AI identities introduce new risks, requiring governance that observes execution paths rather than relying solely on static policy definitions.

• Identity Fabrics address credential sprawl in hybrid clouds by connecting policy intent with runtime behavior
• Non-human identities require lifecycle governance comparable to human accounts to reduce attack surfaces

Datacenter operationsData Center Knowledge · 2026-08-28

DOE's Transmission Corridor Decision Impacts Data Center Growth

The U.S. Department of Energy's (DOE) recent decision not to designate three proposed National Interest Electric Transmission Corridors (NIETCs) has raised concerns about how utilities and grid planners should prepare for increasing electricity demand, particularly from data centers. The decision came despite DOE's draft 2026 National Transmission Needs Study, which highlighted accelerating demand from data centers, manufacturing, and electrification as major drivers for future transmission capacity.

Utilities face a dilemma: building transmission infrastructure early risks creating unused capacity, but waiting could leave the grid unprepared for projected demand. Texas’s Competitive Renewable Energy Zones (CREZ) program is cited as a successful example of proactive planning, enabling data center growth in areas with robust transmission infrastructure. However, replicating this approach elsewhere requires balancing evidence of future load with the need for timely investment.

Former DOE Secretary Jennifer Granholm criticized the NIETC decision on LinkedIn, stating, 'You can’t declare an energy emergency, demand that America pump out more electricity, and then purposefully make it harder to deliver that power where it’s needed.' The Energy Systems Integration Group (ESIG) recommends proactive, scenario-based planning to address the mismatch between long transmission timelines and rapid data center development.

Despite the challenges, experts suggest a staged approach to transmission planning, securing rights-of-way and scalable designs early while deferring major capital investments until there is stronger evidence of durable load. This strategy aims to minimize stranded assets while ensuring the grid can support future demand.

• DOE's decision not to designate three transmission corridors raises concerns about grid readiness for data center growth.
• Proactive planning, like Texas’s CREZ program, can enable infrastructure for future demand but requires evidence-based investment.

AI in enterpriseCIO.com · 2026-08-28

Unlocking Enterprise Intelligence Through Knowledge Worker Expertise

As artificial intelligence (AI) becomes increasingly integral to enterprise operations, the key to unlocking its full potential lies in empowering knowledge workers. According to Andy MacMillan, CEO of Alteryx, line-of-business employees—such as business analysts, finance leaders, and supply chain managers—possess critical expertise that no AI system can replicate. Their involvement is essential for building and deploying trustworthy AI workflows.

For AI to deliver reliable results, organizations must prioritize transparency, understandability, repeatability, and auditability in their systems. While over 90% of organizations are using AI, only 28% trust it for decision-making, highlighting the importance of trust in operationalizing AI effectively. Trustworthy AI requires clean, reliable data, but even more critical is the application of business logic by knowledge workers to ensure accurate outputs.

The successful integration of AI into enterprise workflows depends on collaboration between IT teams and knowledge workers. While IT can build the tools, it is the domain experts who must define the logic and oversee execution. This partnership ensures that AI systems remain adaptable to regulatory changes and business needs, ultimately enhancing operational efficiency and decision-making.

Enterprise intelligence rests on five core pillars: trustworthy and transparent data, empowered business analysts, shared responsibility across leadership, cross-functional collaboration, and evolving leadership practices. By leveraging the expertise of knowledge workers, organizations can harness AI to drive meaningful insights and superior business outcomes.

• Knowledge workers' expertise is critical for building and deploying trustworthy AI workflows.
• Trustworthy AI requires transparency, understandability, repeatability, and auditability.

AI in enterpriseInfoWorld · 2026-08-28

Why Enterprise AI Projects Continue to Fail

Enterprise AI projects often fail not because of technological limitations but due to organizational readiness and strategic misalignment. Many organizations prioritize deploying AI models without clearly defining the business outcomes they aim to achieve. This leads to projects that generate impressive demos but lack measurable business value.

A common failure pattern is the disconnected pilot, where AI systems function well in controlled environments but struggle to integrate with essential enterprise workflows like ERP, CRM, or supply chain platforms. Scalability requires robust architecture, governance, and operational planning, which are often overlooked during pilot phases.

Another critical issue is poor data governance. Generative AI relies on trusted, well-governed data to produce reliable outputs. Fragmented, duplicated, or stale data can lead to fluent but inaccurate results, amplifying existing data problems rather than solving them.

Additionally, enterprises often misunderstand the economics of AI. While lab-scale projects may appear cost-effective, scaling introduces significant expenses related to tokens, embeddings, infrastructure, and security. Without clear cost-benefit analysis, AI projects can become financially unsustainable.

Finally, governance and security are frequently addressed too late in the process. Enterprise AI systems handle sensitive data, requiring robust compliance frameworks and operational controls from the outset. Without these, projects often stall before reaching production.

• AI projects fail due to organizational readiness, not model weaknesses.
• Poor data governance amplifies risks in generative AI deployments.