Openmind Networks announces the launch of its Campaign Detection and LLM-based Analysis module, a new addition to the OpenProtectAI fraud prevention suite.
DUBLIN, LEINSTER, IRELAND, May 7, 2026 /EINPresswire.com/ — Openmind Networks today announces the launch of its Campaign Detection and LLM-based Analysis module, a new addition to the OpenProtectAI fraud prevention suite. The module applies vector database technology and Large Language Model (LLM) evaluation to identify and block coordinated messaging fraud — including smishing and Artificially Inflated Traffic (AIT) — at the campaign level rather than the individual message level.
Beyond Message Filtering
Traditional fraud firewalls assess messages in isolation. As threat actors increasingly use generative AI to craft varied, low-signature attack patterns, this approach is becoming insufficient. The new module addresses this by grouping messages according to structural and semantic similarities in real time, surfacing coordinated Campaigns that would otherwise evade detection.
By analysing the intent of a messaging stream as a whole, OpenProtectAI assigns a dynamic Threat Score to each identified campaign, enabling operators to automatically block attacks that appear legitimate to conventional filters.
Subscriber privacy is central to the architecture. Prior to any AI analysis, all message content undergoes an automated depersonalisation process:
– Personally Identifiable Information (PII) — including names, phone numbers, and email addresses — is automatically redacted or hashed.
– Original message text is never stored to disk or accessible to analysts.
– Only anonymised, structural data is passed to the LLM for intent evaluation.
Multi-Model LLM Architecture
The module features a model-agnostic interface that seamlessly integrates with a customer’s preferred LLM—including (but not limited to) Google’s Gemini, OpenAI’s ChatGPT, and Anthropic’s Claude. This flexible architecture allows operators to deploy the specific model that best aligns with their functional requirements and local regulatory constraints.
Field Trial Results
In recent field trials using Gemini, the analysis engine achieved a 99.9% accuracy rate in fraud identification across multiple jurisdictions. For one Tier 1 operator, the module was projected to protect against up to €35 million in annual losses from AIT and smishing.
“Operators are facing a new generation of fraud that is deliberately designed to bypass first-generation defences. This module gives operators the analytical depth to identify fraud at the campaign level — without ever processing or retaining raw subscriber data.”
— Valentina Novara, Head of Customer AI & Analytics, Openmind Networks
About Openmind Networks
Openmind Networks is an independent technology company focused on mobile messaging software services for the telecom industry. Openmind Networks’ messaging solutions enable telecom operators to consolidate their core messaging, protect their network and unlock the potential in business messaging and 5G.
Boasting a highly experienced team of engineers, Openmind Networks has consistently led the way in bringing new innovations to the mobile messaging industry for more than two decades.
Openmind Networks is responsible for delivering more than 1.5 billion messages per day with a global customer base which includes the world’s largest mobile operators, wholesalers, aggregators, social media providers and software firms.
Media Contact
Brendan Tobin
Director of Marketing
Openmind Networks
+353 (0)1 633 0070
marketing@openmindnetworks.com
For more information, visit www.openmindnetworks.com or contact +353 1 633 0070.
Brendan Tobin
Openmind Networks
+353 1 633 0070
email us here
Legal Disclaimer:
EIN Presswire provides this news content “as is” without warranty of any kind. We do not accept any responsibility or liability
for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this
article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
Media gallery
