Advisories illustration
BGD e-GOV CIRT published an advisory on 9 July 2026 stating that threat actors are increasingly abusing the branding of popular AI services as social-engineering bait to drive phishing, credential theft, payment fraud and malware delivery. Microsoft, as cited by CIRT, reported campaigns using the names and visual branding of ChatGPT, Claude, DeepSeek, Microsoft Copilot and other AI tools. Microsoft assessed that the observed incidents reflected abuse of trusted brand names as lures rather than compromise of the legitimate AI vendors themselves. CIRT said the activity spans phishing emails with urgent account or billing themes, adversary-in-the-middle token theft, malvertising that pushes fake downloads, and SEO-driven discovery of malicious GitHub repositories. It assessed the threat as directly relevant to Bangladesh because observed activity had already affected South Asia, while the lures depend on curiosity, urgent billing prompts, cloud-hosted redirects and fake software downloads that are equally plausible against Bangladeshi government staff, students, freelancers and mobile-first internet users. Playbooks, CIRT wrote, can be localised to Bangla or English with minimal attacker effort. Four Bangladesh risk scenarios were listed. First, people who use AI services for study, coding, translation, graphics or office work may respond to fake subscription or access-restoration messages. Second, users searching for free or downloadable AI tools, plugins, cracked software or model installers may hit malicious GitHub repositories and malvertising. Third, government officials, university staff, financial-sector employees, outsourcing professionals and software developers may face elevated exposure because they open attachments and use cloud identities. Fourth, widespread use of personal Windows devices in homes, education and small businesses increases the chance that fake installers will be executed outside managed controls. Potential impacts listed included payment-card theft, credential compromise, AiTM session hijacking, infostealer infections such as Vidar, account takeover and reputational damage through misuse of compromised accounts. Microsoft, as summarised by CIRT, linked observed campaigns to collection of payment-card data, theft of credentials and tokens, and delivery of Vidar Stealer, Lumma Stealer, Hijack Loader and Oyster. Part of the malware-distribution activity was linked to Storm-3075, with malware-signing services attributed to Fox Tempest. On 5 May 2026 Microsoft detected a ChatGPT-themed cluster whose subject line warned that ChatGPT Plus would be downgraded within seven days unless payment details were updated; 4,500 emails in one cluster targeted South Africa, with related activity delivering as many as 100,000 emails in a single day. From 20 to 22 April 2026 a Claude-themed campaign targeted more than 2,000 organisations, primarily in the United States, the United Kingdom and India, using an attached “Fill and Sign Claude Appeal Form.pdf”. A malvertising lure for an “Awesome AI Windows Plugin” on 13 March 2026 targeted more than 66,000 devices, delivering a signed installer from a GitHub repository. In April 2026 a fake DeepSeek V4 GitHub organisation hosted malicious archives that installed Vidar. What this means for organisations in Bangladesh: Treat AI-branded billing, appeal or plugin-download messages as high-risk. Block the infrastructure CIRT listed, including dash.awaydouble.org, servicing.pureplantcravings.com and brokeapt.com, and hunt for pythonw.exe created under AppData\Local after GitHub downloads. Report incidents to cirt.gov.bd/report-incident, cti@cirt.gov.bd or info@cirt.gov.bd. Primary source: BGD e-GOV CIRT advisory “AI-Branded Social Engineering, Phishing, and Malware Delivery Campaigns”, 9 July 2026.

Sources

  1. BGD e-GOV CIRT — AI-Branded Social Engineering, Phishing, and Malware Delivery Campaigns2026-09-24

Cyber71 links the primary material behind each report so readers can check it directly. External pages are not under our control and may change.

All news

Need help with a security issue?

Describe the situation and we will open a private case for follow-up.

Get Security Help