Google Cloud says 78% of surveyed leaders see legacy upkeep diverting resources from AI
Newly detailed findings put maintenance workloads alongside vendor dependence. But the survey’s definition of legacy software covers locally running systems—not just old code.
Google Cloud’s survey of 1,250 U.S. enterprise IT and business decision-makers reports two distinct constraints: 78% said maintaining locally run legacy systems diverts resources from AI innovation, and 78% said vendor lock-in limits their ability to choose AI. Another 75% said existing enterprise software limits their ability to use AI fully. VP Jeanette Manfra uses the findings to argue buyers and policymakers should preserve independent choices across software, cloud and AI, while shifting more security responsibility to software makers. The results are self-reported, and Google also recommends its own security products.
01
49% said reliance on on-premises software had delayed, narrowed, paused or canceled AI or other technology initiatives; the measure covers projects beyond AI.
02
59% of IT leaders described the time spent maintaining, patching, integrating and securing legacy enterprise software as unreasonable.
03
39% cited complex licensing or limited negotiating flexibility; 80% said a provider’s installed base should not dictate their AI choice.
Maintaining existing software is competing with AI for money and engineering talent, according to findings Google Cloud executive Jeanette Manfra detailed on November 19. In a Federal News Network commentary, she said 78% of surveyed enterprise technology leaders reported that legacy-system maintenance directly diverts resources from AI innovation—a finding she used to argue for greater vendor choice.
Manfra, Google Cloud’s vice president of global risk and compliance, was elaborating on research the company published earlier this month. Google Cloud and Hypothesis Group surveyed 1,250 U.S. enterprise IT and business decision-makers. The study covers public- and private-sector organizations, connecting complaints about daily maintenance with respondents’ concerns about dependence on established technology providers.
Maintenance burdens are not the same as canceled projects
The study’s terminology matters. It defines legacy technology as software running locally on computers and servers rather than natively in the cloud. That is a distinction about where software runs, not simply how old it is. The findings should therefore be read as respondents’ assessments of that locally running software footprint.
The earlier publication supplied a concrete measure of disruption: 49% said reliance on on-premises software had delayed, narrowed, paused or canceled AI or other technology initiatives. Those outcomes span several levels of disruption, and the question includes technology projects beyond AI.
She said 59% of IT leaders described the time spent maintaining, patching, integrating and securing legacy enterprise software as unreasonable.
Two different constraints on AI
75%Existing software
Said existing enterprise software limits their ability to take full advantage of AI.
78%Vendor lock-in
Agreed vendor lock-in limits organizations’ ability to choose the best AI.
Provider dependence adds a commercial constraint
Maintenance is only one part of Google’s argument. In the study, 39% of decision-makers said reliance on legacy software and cloud providers meant navigating complex licensing or contracts, or limited negotiating flexibility. Unlike the maintenance findings, this measure explicitly includes cloud providers: the reported commercial constraints are not confined to software running on local servers.
Respondents also expressed a preference about future purchasing. Eighty percent agreed that a provider’s installed base should not determine which AI technology an organization uses.
Security concerns reinforce that preference. Manfra said 76% of technology leaders viewed dependence on a single vendor as increasing systemic cybersecurity risk. Google’s accompanying argument is that concentrating operating systems, productivity applications, cloud hosting and AI models within one provider creates a single point of failure.
Google’s prescription goes beyond migration
Manfra calls for changes to purchasing and software security, not just a different hosting location. Her recommendations ask enterprise buyers and policymakers to preserve choice, while placing more responsibility for security on the companies building software:
Allow organizations to choose software, cloud and AI independently of their existing technology suppliers.
Require software makers to build security into products rather than leaving customers with continuing repair work.
Use AI to find and fix vulnerabilities in software that is already deployed.
Google is also selling tools into the problem it describes. Its research post recommends Security Command Center and Mandiant threat analytics to identify attack paths, prioritize exposure and automate remediation across hybrid environments.
Manfra’s purchasing argument leaves room for organizations to keep their existing systems. Her stated demand is that staying should be a choice—not a condition imposed by the software an organization already runs.
Sources
federalnewsnetwork.comWhen the legacy software meant to delivery AI becomes a trap - Federal News Network
cloud.google.comNew survey finds enterprises trapped between legacy software and AI modernization | Google Cloud Blog
Reader comments
Newest comments first. Replies stay oldest first.