New service offering “Trusted AI Chatbots & Assistants”

diconium also ensures trust in chatbots and assistants as the key to AI success

Stuttgart, June 12, 2024

diconium, the partner for digital growth along the entire value chain, is now launching its new “Trusted AI Chatbots & Assistants” service offering. It provides companies with valuable support in the development of trustworthy AI chatbots and assistants - from the initial analysis to consulting and implementation. The aim of the new service offering is to create tailor-made AI solutions that are protected against hallucinations, the disclosure of internal information and other risks. To this end, diconium relies on a holistic development approach that enables risks to be dealt with systematically based on a new type of quality and risk taxonomy in the four categories of quality, safety, security and law. 

David Blumenthal-Barby, Principal Specialist AI bei diconium: "Building an AI chatbot or assistant in 20 minutes is now entirely possible - but definitely not a good idea. This is because off-the-shelf AI chatbots can entail risks in terms of quality, safety, security and law. This has recently been demonstrated again by a number of practical examples. In order to create trust in AI bots and assistants, a systematic approach is required to minimize the risks. In other words, trust is also the key to AI success with chatbots and assistants. With 'Trusted AI Chatbots & Assistants', we are offering precisely this opportunity and further expanding our position as a trustworthy partner in the field of artificial intelligence." 

Developing trustworthy AI chatbots requires systematic handling of risks

AI chatbots and assistants offer companies a wide range of internal and external applications for more efficient processes and increased customer and employee satisfaction. However, incorrectly programmed AI chatbots and assistants can have the exact opposite effect and, among other things, provide incorrect information, disclose internal information, behave inappropriately or expose the operator to legal risks. To prevent this, diconium provides support with a holistic development approach. With a focus on four key risk areas: 

• Quality: Successfully introduced AI chatbots and assistants have practical added value for their users. This is the only way to ensure a positive return on investment for companies. 

• Safety: Users and operators are protected against misconduct by the AI chatbot or assistant (e.g. incorrect information, damage to reputation). 

• Security: High security standards protect the AI chatbot or assistant against attacks - especially against new attack patterns that specifically target AI systems. 

• Legal: Legal risks (e.g. data protection) of the company- and industry-specific use of AI chatbots and assistants are taken into account from the outset.