According to Sophos new report called Beyond the Hype: The Realities of AI for Cybersecurity, even though 65% of organizations are using generative artificial intelligence (GenAI) tools, 89% of IT leaders are concerned that flaws in these tools could put their organization at risk.
Another report from Sophos X-Ops, titled “Cybercriminals Still Not Fully on Board the AI Train (Yet),” shows that some criminals are starting to use AI. They found that, while many criminals are still unsure about GenAI, some are using it to help with simple tasks like sending a lot of emails and analyzing data. Others are adding it to their spam and social engineering tools.
Almost all organizations (98%) have some form of AI in their cybersecurity systems. However, IT leaders are worried about depending too much on AI. About 87% of them are concerned that this could lead to a lack of responsibility when it comes to cybersecurity.
Larger organizations (over 1,000 employees) focus on better protection. In contrast, smaller companies (with 50-99 employees) value reducing stress for their workers the most. Despite this, 84% of leaders worry that there will be pressure to cut cybersecurity jobs because people think AI can do everything that humans can do.
Other Important Points from the “Beyond the Hype” Report: GenAI Costs Are Difficult to Measure: 75% of IT leaders say it is hard to measure how much GenAI costs in cybersecurity products. Companies Hope to Save Money with GenAI: While 80% of IT leaders believe that using GenAI will greatly increase the cost of cybersecurity tools, most think that it can help reduce overall spending. About 87% believe that the savings from using GenAI will make up for its costs.
“As with many other things in life, the mantra should be ‘trust but verify’ regarding generative AI tools. We have not actually taught the machines to think; we have simply provided them the context to speed up the processing of large quantities of data,” said Chester Wisniewski, director, global field CTO, Sophos. “The potential of these tools to accelerate security workloads is amazing, but it still requires the context and comprehension of their human overseers for this benefit to be realized.”
