Survey Surfaces Increased Cybersecurity Risks Following AI Adoption
Survey Surfaces Increased Cybersecurity Risks Following AI Adoption
Publish Date: 2026-02-25 09:54:00
Source Domain: securityboulevard.com
Using an unordered list, summarize the following article with between 4 and 8 key points.
A global survey of 2,000 IT decision makers published today shows cybersecurity risks are rising as more organizations embrace artificial intelligence (AI) applications.Conducted by Sapio Research on behalf of Fastly, the survey finds that cybersecurity incidents impacting organizations that have integrated AI into core processes incur 135% more costs than organizations that have not.Additionally, three quarters of organizations (75%) that have invested in AI require an average of 6.8 months to recover from security incidents, a full 80 days longer than other organizations.Fernando Medrano, deputy CISO for Fastly, said the survey makes it clear that AI has become a significant attack vector that adversaries are looking to exploit. In fact, more than a third (34%) of respondents that work for organizations that have embraced AI said direct exploitation of those tools and applications contributed to their last incident, with another 30% saying that AI use led to an oversight that contributed to the incident.Unfortunately, the survey also makes it clear that many cybersecurity teams lack the expertise needed to effectively respond to these types of incidents. More than half (53%) of respondents said their teams lack AI-specific expertise to address emerging threats. A similar percentage (51%) of organizations that have embraced AI also report confusion over who handles incident response.Overall, 90% of respondents report their organization experienced at least one cybersecurity incident. On the plus side, the number of security incidents reported year over year stayed flat. Organizations faced an average of 41 known incidents, with software bugs triggering 40% of those incidents, followed by external attacks (39%) and misconfigurations (25%). However, two thirds (66%) of respondents report their organizations suffered repeat occurrence within three months of an initial incident.Over half (52%) also report investing in post-incident reviews to systematically analyze what went wrong and how to prevent it next time. Another 43% have implemented response automation, but 30% of organizations still lack regularly tested incident response playbooks.Cybersecurity teams are also starting to better understand the scope of the attack surface that needs to be defended in the age of AI, said Medrano. In addition to investing in tools to discover AI agents (56%), survey respondents also report their organizations are investing more in securing application programming interfaces (APIs) and deploying web application firewalls (WAFs), the survey finds.In many instances, it is more cost-effective to better secure the backend services being invoked using AI tools than it is to try to secure every tool or application that has been deployed, noted MedranoOn the down side, survey respondents also report infrastructure costs are rising as AI bots scrape content from their websites. Approximately two thirds of businesses (64%) say AI scraping has become a material cost center, with expenses soaring over $348,000 annually on average. Another 40% report operational disruption, while 29% are dealing with user experience problems such as sluggish load times, broken functionality, and degraded performance.It’s obviously still early days so far as AI adoption is concerned, but in the short term at least, it would appear cybersecurity teams may want to prepare for the worst while hoping for the best.
