Data in Action

Technology professionals say data privacy tops their list of ethical worries surrounding the deployment of generative AI in the enterprise, according to a Deloitte report. The firm surveyed 1,848 business and technology professionals.

Nearly three-quarters of professionals ranked data privacy among their top three ethical concerns about the technology's use, according to the report. Two in 5 respondents flagged data privacy as their No. 1 concern this year, almost double the 1 in 4 that cited data privacy in Deloitte's 2023 survey. 

Tech leaders are poring over the infrastructure and talent needs of their organizations as they help guide generative AI adoption. Ethical concerns should also make it on the checklist.

"GenAI collapses the ‘expertise barrier’: more people can get more out of data, with less technical knowledge needed," said Sachin Kulkarni, managing director, risk and brand protection, Deloitte LLP, in the report. "While a benefit, the potential for data leakage may also increase as a result."

Professionals are also worried about the impacts of generative AI on transparency, data provenance, intellectual property ownership and hallucinations. Job displacement, though often cited as a top concern, was only flagged by 16% of respondents.

Across emerging technology categories, business and IT professionals identified cognitive technologies — a category that includes large language models, machine learning, neural networks and generative AI, among others — as posing the most severe ethical risks. 

The category surpassed digital reality, autonomous vehicles and robotics, among other technology verticals. However, respondents also ranked cognitive technologies as the most likely to drive social good. 

Due to its reliance on data, the majority of executives are concerned about how generative AI tools can increase cybersecurity risks by expanding their attack surface, a Flexential survey published earlier this month found.

In the race to extract value from large language model technologies, enterprises have had to take a long, hard look at their data estates and ask difficult questions about foundational infrastructure.

For many companies, the path to clean analytics, data security and generative AI ROI runs directly through vast stores of customer data located across various business units.

USAA began a massive data estate modernization roughly three years ago, when the century-old banking and insurance giant tapped Ramnik Bajaj as SVP and chief data and analytics officer.

The end goal is a “unified data environment,” according to Bajaj, who spent three years leading data operations at Wells Fargo and two decades as a partner in Deloitte’s financial services data practice prior to his current role.

“There were multiple warehouse and analytics environments on-prem and our journey was to get all of that over to the cloud,” Bajaj told CIO Dive.

In a unified, cloud-based environment, stakeholders can perform analytics without copying data or creating additional repositories. “The data is already there with the analytics tools,” Bajaj said.

Enterprisewide data transformation takes an upfront commitment of time and resources. The price tag for data estate modernization will average $35.5 million this year, according to a recent IT executive survey by cloud database provider Couchbase.

For data-dependent insurance and banking companies, the costs — and the stakes — are even higher.

Citi just absorbed $135.6 million in regulatory fines for data deficiencies after what CEO Jane Fraser characterized as “decades of underinvestment” in the bank's risk and control technologies earlier this month. Bank of America poured billions into modernizing its data systems over the last decade, CEO Brian Moynihan said during a January earnings call.

Migration motivation

Keeping a yearslong migration journey on track can be a challenge. In addition to clearing tangled webs of technical debt, IT teams have to consider potential operational disruptions and inconveniences to end users during the switch.

Bajaj used a “carrot-and-stick” approach to sustain modernization momentum, setting firm migration deadlines while rolling out better reporting tools. A data and analytics organization within USAA oversaw strategy, defined governance policies and standards and set annual migration priorities.

“We had a sunset date for each environment,” Bajaj said. “The sunset date was the stick and the carrot was the analytics and reporting tools available on the cloud environment.”

To accelerate migrations, Bajaj centralized engineering processes while federating data decisions. Engineers built a universal migration pipeline and toolkit in Snowflake’s data cloud and let business units choose which datasets to move.

“We took the work away from all of these various teams that own these data assets and warehouses and said, ‘you define what you need moved into the Snowflake data cloud and we will write the ETL code to physically create the pipeline,'” Bajaj said.

USAA remains unaffected by the recent Snowflake data breach that impacted AT&T and as many as 100 other customers, the company told CIO Dive Wednesday.

One of USAA’s priorities was to create a core customer data warehouse with centralized access controls and governance.

“We can manage all of the core data about our members out of a central location,” Bajaj said. “We define the policies and standards through which we govern data and also work with all of the business units to implement those standards and requirements.”

As the company settled into Snowflake, its data posture shifted from defense to offense — from standing up governance to reaping analytics rewards.

“Our focus has moved to deploying AI-based solutions and getting more insights from the data,” Bajaj said. “As our data quality matured, we shifted to using those data assets for helping our members.”

Sponsored content

By Vertex

The challenge

IT teams have a long list of responsibilities—and often times limited resources to manage them. Tax compliance is a high priority item on that list—and a way to effectively handle it is by putting data in action through tax automation and moving away from manual tax compliance tasks. From an organizational perspective, this type of shift is essential to overall business success since:

• Challenges with latency during peak processing times can lead to delays in tax-related operations and reporting.
• Providing transparency and traceability using traditional, manual systems can be extremely challenging, time-consuming and prone to errors.
• Different regions have distinct laws and regulations, including data privacy, cybersecurity and taxation rules. It can be challenging for IT departments to ensure compliance across all regions in which they operate.
• Discrepancies or errors in data can lead to inaccuracies in tax calculations.
• Maintaining consistency in tax data between different systems (e.g., ERP, CRM and tax software) is challenging. Inaccurate data synchronization can lead to discrepancies and compliance risks.

The potential of digital tools for tax

These are just a few of the reasons why IT teams aim to leverage the full potential of digital tools to effectively manage data to optimize their tax-related operations and decision-making—while also helping ensure accuracy and compliance. If this type of digital approach isn’t implemented, the business may miss optimization opportunities, hindering a business’s efficiency and growth potential.

“Sales tax automation that utilizes current, verified tax information from a centralized sales tax engine and seamlessly integrates into ERP, e-commerce, or point-of-sale (POS) systems, is more than just a business system — it’s a risk management and financial management tool,” says Alexis Brink, Director of Solutions Marketing for Vertex.

Using a tool makes it possible to potentially minimize tax obligations through improved data quality that aids compliance—and can result in reduced overpayments of U.S. sales and use tax. It’s also possible to increase efficiency with rules and automation. In addition, when IT teams collaborate with tax teams, the solutions they develop can reduce the amount of time spent on the identification, correction and validation of compliance submission data through powerful automation capabilities, including supplied rules for common scenarios including data completeness, logical checks and destination validations—and create rules that are specific to a particular business.

Improved data accuracy and completeness with powerful validation capabilities also means that with increased efficiency there is the ability to easily review errors, warnings, missing or incomplete info (based on your rules), create your own rules to correct, augment, or route data and finalize data for the next step in your compliance process.

Effective data tools can also help lower audit exposure and risk of penalties. Minimizing risk and exposure through the automation and standardization of data validation and correction process can provide a valuable audit trail for audit defense.  For example, an audit history log provides full transparency into all data repairs made; automatically captures and documents all comments and data changes; tracks application activities by user; and minimizes the time required to get backup information for specific transactions.

The power of tax and IT teams collaborations

By working together, the tax and IT teams can develop solutions that utilize and maximize the strengths of both teams—benefiting the organization as a whole. The enhancements impact the following areas:

• Tax compliance: IT’s expertise with enterprise resource planning (ERP) systems, as well as other tools used by finance departments, is invaluable. IT departments often have the budget for tools that update the ERP systems, so it’s important that the two teams work closely together.
• Tax planning: Machine learning technologies can play an important role in helping to ensure correct tax categorization for new products and services, across new geographies.
• Audit defense: Auditors’ initial focus isn’t always on transaction amounts – they may be more interested in exemptions, credits and adjustments. IT can help ensure that all the relevant tax data is clean, safely stored and quickly accessible. 

The benefits of the cloud

When tax technology tools are hosted in the cloud, it allows for greater scalability, cost-efficiency, agility, security and automated maintenance. It also provides access to cutting-edge technologies, supports global reach and allows IT to focus on more strategic initiatives and resource optimization. 

Conclusion

When IT teams establish robust collaborations with tax teams, the result is effective use of data, an evolution away from time-consuming, manual processes to tax automation—providing the tools that enhance compliance and increase efficiency through accurate tax determination; invoice generation, digital reporting and archiving; analytics; and reporting transparency. An even greater benefit generated by IT taking a proactive approach to tax automation and data management is the impact on team members—freeing up time to focus on strategic tasks that help the business thrive and grow.

Executives are targeting upskilling investments toward cyber, AI and data training opportunities as hiring for these skills remains a persistent pain point, according to a Skillsoft report. The corporate training company surveyed 5,711 IT decision-makers and staff, including 219 C-level executives. 

More than one-third of C-suite leaders identified cybersecurity and AI as top investment areas for training, and 3 in 10 said data. A year ago, only 13% of executives said hiring for AI skills was difficult — 32% said finding AI talent was challenging this year. 

“The reality is you can’t have a department operating at 100% efficiency when you don’t have a workforce qualified to meet those expectations,” Skillsoft said in the report. “And, you can’t take full advantage of emerging technologies without skilled staff.”

AI, data and cybersecurity have always been intertwined, but the need for highly skilled teams in these areas is reaching a boiling point.

Without a good foundation and understanding, teams are left to toil. 

“Gaps in skills necessary to keep a department running seamlessly is a heavy weight on IT leaders’ shoulders,” Skillsoft said in the report. “Decision-makers need to get creative with how they handle skilling, reskilling, and upskilling — whether it’s addressing root causes behind recruitment or retention issues or investing in more effective, efficient training solutions.”

Businesses have incentives to plug the holes. Critical gaps in skills can put pressure on those ultimately bearing the burden: existing IT staff. Employees who feel overworked and behind on projects aren’t having a great experience at work, which can lead to them searching for greener pastures. 

Women in tech plan to leave their jobs this year at a comparable rate to that of 2022’s "great resignation" era of high attrition, according to an Ensono report. Enterprises that aren’t willing to support employees at work are putting their aspirations at risk. 

“You already have a known commodity: a workforce that knows your business, your customers, your process,” Skillsoft said in the report. “What you might not have, in terms of new skills or expertise in emerging technology, can be developed by dedicating resources to new certifications and continuous learning.”

Most enterprises are leaning on their mainframes to run generative AI workloads, according to a Kyndryl report. The IT services provider commissioned Coleman Parkes Research to survey 500 business and technology executives for its second annual mainframe modernization report.

Nearly 9 in 10 respondents plan to deploy generative AI in mainframe environments as part of a hybrid strategy that leverages public cloud while modernizing on-prem infrastructure. While 86% of respondents said mainframes remain essential, 96% plan to migrate a portion of their applications to cloud.

“Mainframes continue to occupy a central role in the hybrid world and are evolving to serve new use cases, with AI and security increasingly influencing modernization plans,” Petra Goude, global practice leader for core enterprise and zCloud at Kyndryl, said in a release accompanying the report.

Large language models thrive in massive hyperscaler data centers, where AI-optimized chips and compute resources are relatively easy to find. But many organizations are finding a home for smaller, purpose-built generative AI tools on-prem, aboard mainframes that house core data and business applications.

To ease on-prem AI adoption, IBM rolled out high-capacity mainframe processors for its zSystems and LinuxOne clients last month. Dell is also banking on enterprises running AI on-prem as fuel for its GPU-infused server and workstation business, the company said earlier this month.

The option of bringing the model to the data rather than sending the data to cloud is a potential solution to AI security concerns, which half of the survey’s respondents cited as one of the main motivators for mainframe modernization.

“You can't do machine learning and cool AI stuff without good data,” Brian Klingbeil, chief strategy officer at IT services company Ensono, told CIO Dive. “And if you have a mainframe, what kind of data is on that mainframe? The answer almost 100% of the time is: The most important data to my company — and it's there in spades.”

Despite a rush to cloud for many customer-facing applications, organizations report running 56% of critical workloads on a mainframe, according to the Kyndryl survey. More than half of respondents said mainframe usage increased this year and 49% expect that trend to continue.

“Cloud is great for systems of customer engagement, like a mobile check-in app for an airline,” Klingbeil said. “But mainframes tend to be systems of record for airline flight scheduling, insurance claims processing, ATM machines, stock ledgers and large retailers.”

Talent woes continue to plague mainframe modernization efforts, as experienced practitioners retire and are replaced by technologists primarily trained in the cloud. More than one-quarter of respondents expressed concern that their organization may lack the necessary mainframe modernization skills and slightly fewer said they’ve already encountered talent shortfalls, Kyndryl found.

Lack of familiarity with mainframes can also be an executive-level issue, Klingbeil said.

“Some younger CIOs, when they take over, don't really realize that it supports massive revenue streams, even if it’s a big line item. If you do the math and look at dollar-for-dollar revenue, mainframes are relatively efficient.”

Legacy tech upgrades cost the average business $2.9 million in 2023, a SnapLogic survey of 750 IT decision-makers found. Nearly two-thirds of businesses invest more than $2 million on maintaining and upgrading legacy systems, according to the report.

More than 3 in 5 IT leaders said their company’s data stack is experiencing a moderate to severe negative impact due to technical debt, including outdated code.  

Despite ambitious goals to innovate and streamline operations, businesses face significant roadblocks when modernizing legacy technology systems. 

Challenges are inherent in major technology upgrades, especially for legacy systems, according to Jeremiah Stone, CTO of SnapLogic. 

More than three-quarters of IT decision-makers said their teams spend 5 to 25 hours a week updating and patching legacy systems, according to the survey. Legacy tech can also sap productivity and impact the bottom line. 

Maintenance and development costs continue after the initial implementation phase, Stone said. Companies also struggle to dissolve technical debt when legacy systems are intertwined with existing workloads, making them imperative to everyday operations. 

Legacy tech is putting emerging tech ambitions at risk as well. Nearly one-third of respondents said up to 25% of their legacy systems are unable to support AI tools and workloads, according to the survey. 

“This is a time for CIOs and CTOs to not only be creative on how they do more with less while building out their GenAI initiatives and projects but also leapfrog the competition with calculated investments,” Stone said. Executives should prioritize projects that reduce technical debt, create new revenue streams or democratize technologies, according to Stone.

Walmart is using generative AI to improve the data quality of its product catalog in pursuit of CX benefits for its customers and associates, executives said on a Q2 2025 earnings call.

Cleaner data can help associates find specific products by quickly accessing images of packaging. Customers using Walmart’s AI assistant will soon receive follow-up answers to their queries that will help them find the right item, according to Walmart CEO Doug McMillon.

“We're finding tangible ways to leverage generative AI to improve the customer, member and associate experience,” McMillon said during the call.

Walmart executives entered this year wanting to turn their AI aspirations from 2023 into a reality. Now, better data crunching capabilities are improving behind-the-scenes operations and end-user experiences.

“The quality of the data in our catalog affects nearly everything we do, from helping customers find and buy what they're looking for, to how we store inventory in the network, to delivering orders,” John Furner, president and CEO of Walmart U.S., said during the call.

Walmart used large language models to create or improve more than 850 million pieces of data across its product catalog. Without generative AI, the process would have required 100 times the head count to complete in the same amount of time, executives said. 

The primary in-store improvement is that associates’ can use in-store technology like mobile tools to quickly locate inventory and get items on shelves or to waiting customers — a significant upgrade from the “treasure hunt” of finding items in years past, according to Furner.

The company also bolstered internal generative AI adoption in January, when it rolled out a proprietary generative AI tool to an additional 25,000 employees raising the total access count to 75,000 workers. 

Enterprises want to level hurdles that stand in the way of generative AI ambitions, starting with security, costs and data, according to Enterprise Strategy Group research sponsored by Hitachi Vantara. 

More than 3 in 5 organizations have significant gaps in AI readiness when it comes to infrastructure and data ecosystems, the survey of 800 IT decision-makers found. Nearly three-quarters agreed their infrastructure needs modernization before pursuing generative AI projects.

Organizations are bullish on what AI can do for their operations long-term, but most aren’t ready for widescale adoption just yet.

More than half of AI decision-makers are worried about IT teams’ keeping up with the pace of innovation driven by generative AI, according to the Hitachi Vantara report. Enterprises have a lengthy list of to-do’s, from improving data processes and strengthening infrastructure to adapting risk mitigation frameworks. 

Even so, nearly two-thirds say they have identified at least one use case for generative AI. 

“Enterprises are clearly jumping on the GenAI bandwagon, which is not surprising, but it’s also clear that the foundation for successful GenAI is not yet fully built to fit the purpose and its full potential cannot be realized,” said Ayman Abouelwafa, CTO at Hitachi Vantara, in the report. 

For tech leaders and organizations that are moving forward on projects, measuring success isn’t uniform. 

Around 2 in 5 enterprises track progress based on a qualitative impact analysis, accuracy of generative AI responses or user/process quantitative benefits. Slightly less — 38% — are looking at cost savings as a success marker. 

The push for ROI is expected to grow as AI initiatives command a larger percentage of enterprise budgets.