Companies worldwide are increasingly forbidding staff from using public-facing artificial intelligence (AI) services because of the perceived cybersecurity risk.
New research from BlackBerry has revealed that 75 percent of organizations worldwide are either implementing bans or considering doing so. Even worse news for services such as Microsoft-backed ChatGPT is that 61% of the 2,000 IT decision-makers in the US, UK, Canada, France, Germany, Australia, and Japan said that they intend the ban to become long-term or permanent. Their main reasons for doing so are concerns regarding data security (67 percent) and potential reputational damage (57 percent).
But Blackberry adds that, despite their inclination towards blocking widespread use of the burgeoning technology, most IT decision-makers who responded to the BlackBerry survey also believe that professionally managed generative AI applications can have a positive impact on their business. The potential advantages they foresee include increased efficiency (55 percent), innovation (52 percent), and enhanced creativity (51 percent). According to BlackBerry, the majority (81 percent) of the respondents also heavily favor using AI tools for cybersecurity defense. While BlackBerry advises extreme caution regarding consumer-facing generative AI tools, it also recommends that organizations focus on enterprise-grade generative AI, which, properly used, can offer tremendous benefits.
“Banning generative AI applications in the workplace can mean a wealth of potential business benefits are quashed,” says BlackBerry Chief Technology Officer Shishir Singh. “We are also exercising caution with unsecured consumer-grade generative AI tools. As platforms mature and regulations take effect, flexibility could be introduced into organizational policies.”
Nevertheless, BlackBerry has issued a stark series of warnings regarding the unsupervised use of public-facing AI. If sensitive third-party or internal company information is entered into ChatGPT, that information automatically becomes a part of the chatbot’s data model, which can be shared with others who ask relevant questions, resulting in data leakage. BlackBerry also highlights copyright concerns if ChatGPT is used to generate written material inspired by copyrighted property. The same could apply to new software that has been in part developed using a service such as ChatGPT. Terms of service also indicate that ChatGPT cannot be used in the development of any other AI. The use of ChatGPT in this way could easily jeopardize the company’s future AI development.
BlackBerry’s research also revealed that companies are becoming increasingly wary of staff using their smartphones and other devices to access other external online software services outside the corporate IT environment. Over 80% of those surveyed have voiced additional concerns that unsecured apps pose a cybersecurity threat to their corporate IT environment.
