GoKawiilResearchers Detail 'Dark Sourcery' Campaign Poisoning ChatGPT, Gemini With Fake Support Info
Vigilance Security researchers say attackers are seeding tens of thousands of web pages, PDFs, reviews and fake support portals with fraudulent phone numbers, emails and login links, aiming to get AI chatbots like ChatGPT, Google Gemini and Google AI Overview to surface them as trustworthy answers. The campaign, named Dark Sourcery, has reportedly affected at least 374 companies, including Fortune 100 firms, airlines, banks, travel companies and software providers.
GoKawiil's interpretation of the reporting above, not reported fact.
Vigilance suggests attackers may be exploiting AI chatbots' tendency to trust content hosted on high-authority domains such as universities and government sites, which could make this style of manipulation harder to filter than traditional search spam. If chatbots keep surfacing fake contact details as fact, brands could suffer reputational harm and customers could be funneled toward phishing pages without any breach of the companies' own systems.
- Attackers are using SEO-style tactics to inject fake data into sources AI chatbots draw answers from.
- At least 374 companies across industries including banking, airlines and software have reportedly been targeted.
- The technique may exploit chatbots' trust in high-authority domains, complicating detection and mitigation.
Source: darkreading.com — Elizabeth Montalbano, 2026-09-23
Published there as: “Attackers Manipulate AI Chatbots in Mass Disinformation, Phishing Campaign”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.