Despite various government initiatives and increased awareness, coupled with advances in technology, cybersecurity remains an ever-present concern for the modern corporation. These concerns are not going anywhere as 2023 unfolds.
This is the take of Christoph Börner, Senior Director Digital at Cyara. Börner is predicting that 2023 will see at least one high-profile security breach caused by a chatbot. This is in the context of companies using more self-service and web chat customer support options in 2023 due to a combination of consumer preferences and a desire to lower costs.
This could arise, says Börner since: “The same vulnerabilities, threats and risks apply to chatbots as they do to other customer-facing online applications. Chatbot security vulnerabilities can include impersonating employees, ransomware and malware, phishing and bot repurposing.”
There is a common pattern, says Börner: “Many companies that use chatbots don’t have the proper security testing to proactively identify these issues before it’s too late. It is highly likely there will be at least one high-profile security breach due to a chatbot vulnerability next year, so chatbot data privacy and security concerns should not be overlooked by organizations. Threats that are not addressed can result in data theft and modification, which is extremely harmful to a brand, its reputation and its customers. More companies must have chatbot security protocols in place to reduce risk so chatbots can be used with confidence.”
A further weakness with chatbots is reliability and the associated importance of rigorous testing. Börner expresses this weakness as businesses become increasingly reliant upon chatbots, there is the probability that 100 percent of bots will fail without the appropriate level of testing.
According to Börner: “In recent years, the use of conversation artificial intelligence has increased dramatically and since 2019, they’ve increased as a brand communication channel by 92 percent.”
This leads many “Companies are now buying chatbot implementations from big providers like IBM and assume that they are always working correctly. But chatbot development and testing isn’t something you can do once and forget about.”
However, as Börner points out: “To be effective, it must be an ongoing process where the company is adopting new features to meet clients’ needs. This means that the development process is not complete without testing and ongoing monitoring. Even with the proper testing in place, it’s likely that all bots will have technical failures at some point, which if not immediately addressed, leads to unsatisfied customers.”
Therefore, a process of continuous improvement is needed. As Börner explains: “By continuously testing, companies reduce the amount and the severity of chatbot failures to prevent customers from experiencing those problems first-hand.”
