SecurityWeek\u2019s weekly cybersecurity news roundup offers a concise yet comprehensive overview of vital developments that may not receive full standalone coverage yet remain fundamentally relevant to the broader threat landscape. This curated summary highlights key stories across vulnerability disclosures, emerging attack methods, policy updates, industry reports, and other noteworthy events to help security professionals, system administrators, and organizations stay well-informed about the rapidly evolving global cybersecurity environment. <\/p>\n
As artificial intelligence infrastructure continues to experience widespread adoption, threat actors are increasingly adapting their operational methods to target exposed AI and open-source services. Black Lotus Labs has detailed a novel threat dubbed PoeLLM, a malware strain active since at least April 2026. The campaign specifically targets vulnerable services such as LiteLLM, Ollama, Gotenberg, and Gitea with the dual objective of illicitly mining cryptocurrency and expanding a growing botnet. <\/p>\n
What sets PoeLLM apart from traditional botnets is its unique command and control (C&C) architecture. Infected machines are programmed to extract four specific keywords from a poem hosted publicly on GitHub. These words are then dynamically converted into the IP address of the active C&C server, allowing the threat actor to seamlessly switch servers simply by updating the words within the poem. Security telemetry shows that the operator, assessed with moderate confidence to be an Italian-speaking individual or group, has modified the poem 11 times to shift infrastructure and evade detection.<\/p>\n
Software supply chain security remains a critical concern for development teams worldwide, particularly as automated workflows are increasingly leveraged as vectors for compromise. According to findings from GitGuardian, the GhostAction supply chain campaign actively pushed its malicious, secret-stealing GitHub Actions workflow to 772 additional public repositories. These repositories belong to 373 distinct users and organizations, and the campaign targeted an estimated 2,577 sensitive secrets, including SSH keys along with credentials for Amazon Web Services (AWS), Microsoft Azure, and various database systems. <\/p>\n
The activity was observed between August 31 and September 30. Aside from routing stolen data to a newly established exfiltration server, the threat actor largely reused the core playbook deployed in previous operational cycles. GitGuardian\u2019s collected data indicates that this persistent campaign never truly stopped, highlighting the continuous threat posed by automated credential harvesting in modern software development pipelines.<\/p>\n
In the legal arena, accountability caught up with perpetrators of decentralized finance exploits. A federal jury has convicted 36-year-old Jonathan Spalletta of Maryland on multiple counts of computer fraud and money laundering connected to two major 2021 heists targeting the decentralized cryptocurrency exchange Uranium Finance. Spalletta abused underlying smart contract flaws to siphon approximately $1.4 million in the initial attack, followed by a staggering $53.3 million in a secondary assault that ultimately forced the exchange to completely shut down operations. <\/p>\n
Following the thefts, the perpetrator attempted to obscure the origin of the funds by laundering them through a series of complex cryptocurrency transactions, which included the use of privacy-enhancing mixing services like Tornado Cash. Rather than liquidating the assets into traditional fiat currency, Spalletta used the stolen capital to purchase high-value collectibles, including rare Magic: The Gathering and Pok\u00e9mon trading cards, as well as antique Roman coins. Following the conviction, he faces up to 10 years in federal prison for the fraud count and an additional 20 years for money laundering.<\/p>\n
In policy developments affecting federal cybersecurity personnel, the Cybersecurity and Infrastructure Security Agency (CISA) has announced plans to maintain its Cybersecurity Retention Incentive program\u2014which provides bonuses worth up to 25% of base salary\u2014through fiscal year 2027. However, the agency is tightening the criteria required to qualify for the incentive. Under the overhauled guidelines, staff members must now hold an "exceeds expectations" performance rating or higher and spend at least 51% of their working hours performing direct cyber duties within one of three designated job series. <\/p>\n
Employees who fall outside those specific job series but spend 75% or more of their time on cybersecurity responsibilities will still be permitted to apply through a specialized review board. This strategic policy overhaul follows a critical finding by the Department of Homeland Security (DHS) inspector general, which revealed that the previous iteration of the retention program had been mismanaged and applied far too broadly across the agency.<\/p>\n
To address mounting physical and digital convergence risks, the Operational Technology Cybersecurity Coalition (OTCC) has published a detailed proposal outlining what a CISA Binding Operational Directive (BOD) focused exclusively on operational technology (OT) across federal civilian agencies should require. Civilian agencies currently rely on more than 8,000 General Services Administration (GSA)-managed facilities that incorporate complex industrial control networks, including heating, ventilation, and air conditioning (HVAC) systems, power management infrastructure, access control mechanisms, and advanced building automation systems. <\/p>\n
To secure these critical facilities against potential cyber-physical disruptions, the coalition recommends that the directive mandate federal agencies to designate a specific senior official or dedicated office with direct accountability for OT security. Furthermore, the proposal urges agencies to rigorously apply existing baseline requirements and prioritize the implementation of CISA\u2019s core Cybersecurity Performance Goals.<\/p>\n
Consumer-facing brands continue to grapple with automated credential stuffing campaigns that exploit poor password hygiene among users. Fast-food giant Domino\u2019s has begun notifying a small subset of customers that unauthorized third parties managed to gain access to their online accounts. The threat actors leveraged email and password pairs that had been previously leaked in entirely unrelated third-party data breaches. <\/p>\n
Domino\u2019s has emphasized that its core corporate and ordering systems were not compromised during the incidents and that the platform does not store sensitive payment card details on user profiles. Nevertheless, as a precautionary security measure, the company has forced password resets on all affected accounts to prevent further unauthorized access.<\/p>\n
Artificial intelligence is facing increased scrutiny in connection with sophisticated international cyber operations. South Korean President Lee Jae Myung publicly stated that investigators have found strong indications that AI technologies were utilized in a wave of recent cyberattacks targeting the nation’s financial institutions. The sophisticated incidents resulted in the breach of sensitive personal information belonging to bank customers, prompting local law enforcement to launch a full-scale criminal investigation. <\/p>\n
While authorities have not yet publicly disclosed the specific AI tools leveraged or the complete scope of the compromised data, cybersecurity firm CrowdStrike released findings indicating that analysts discovered Claude Code session histories, ARTEX configuration files, and Claude memory files while analyzing infrastructure tied to the attacks. Based on these technical artifacts, CrowdStrike assesses with moderate confidence that a Chinese-speaking, financially motivated threat actor group is likely responsible for the banking campaigns.<\/p>\n
Federal law enforcement authorities have secured a lengthy prison sentence for a key figure behind one of the dark web’s most notorious illicit marketplaces. Raheim Hamilton, a 30-year-old resident of Virginia, has been sentenced to 40 years in federal prison and ordered to pay a $5 million financial penalty. Hamilton previously pleaded guilty to a drug conspiracy charge stemming from his role as co-creator and operator of Empire Market, which he ran alongside Thomas Pavey between 2018 and 2020. <\/p>\n
During its operational lifespan, the dark web marketplace facilitated more than four million distinct transactions totaling over $430 million in illicit revenue. While the platform was primarily utilized for large-scale illegal drug sales, it also functioned as a clearinghouse for stolen credentials, personal identifying information, counterfeit currency, and specialized hacking tools. Pavey, who entered a guilty plea late last year, is scheduled to receive his federal sentencing later this month.<\/p>\n
Hardware monitoring tools deployed within high-performance computing environments have become lucrative targets for attackers seeking to disrupt intensive workloads. Researchers recently disclosed CVE-2026-47483, a high-severity vulnerability residing within Nvidia\u2019s DCGM Exporter GPU monitoring tool. The security flaw allows unauthenticated, remote attackers to flood the tool’s profiling endpoints with malicious requests, triggering severe resource exhaustion and causing the monitoring service to crash. This disruption can severely degrade or halt complex artificial intelligence workloads running on the affected host. <\/p>\n
During global internet scans conducted between March and May 2026, researchers discovered approximately 2,100 distinct hosts exposing the vulnerable DCGM Exporter directly to the public internet without any authentication barriers. This misconfiguration inadvertently leaked detailed telemetry data originating from more than 12,000 high-performance GPUs. Nvidia has since released security updates addressing the vulnerability, and system administrators are strongly urged to update installations to version 4.8.2 or later.<\/p>\n
Software supply chain threats continue to plague software package repositories, with malicious actors successfully injecting credential-harvesting routines into popular development libraries. Version 0.5.144 of According to threat analysis reports, the embedded malware is engineered to automatically harvest sensitive credentials stored in local environments, including keys for npm, GitHub, AWS, Kubernetes, and HashiCorp Vault, alongside configuration data for various AI coding tools. Furthermore, the payload possesses the capability to execute arbitrary code supplied remotely by the attacker and automatically republish compromised versions through other npm packages that the victim’s developer account has publishing rights to.<\/p>\n","protected":false},"excerpt":{"rendered":" SecurityWeek\u2019s weekly cybersecurity news roundup offers a concise yet comprehensive overview of vital developments that…<\/p>\n","protected":false},"author":11,"featured_media":1363,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[313],"tags":[423,323,368,1298,4,143,360,321,1297,1326,320,902,324,1269,674,814,322],"class_list":["post-1364","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-incident-response-forensics","tag-attacks","tag-blockchain-analysis","tag-chain","tag-convictions","tag-crypto","tag-cybersecurity","tag-exchange","tag-forensics","tag-github","tag-heist","tag-incident-response","tag-malware","tag-news","tag-roundup","tag-supply","tag-targeting","tag-tracking"],"_links":{"self":[{"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/posts\/1364","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/users\/11"}],"replies":[{"embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1364"}],"version-history":[{"count":0,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/posts\/1364\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/media\/1363"}],"wp:attachment":[{"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1364"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1364"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1364"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}tensorlake<\/code>, the official npm software development kit (SDK) designed for Tensorlake\u2019s AI agent sandboxes, was maliciously compromised to execute a credential-stealing worm during the package installation process. Security researchers from Socket and Sonatype categorized the incident as a ChainDrop or Shai-Hulud-style supply chain attack. <\/p>\n