Why AI readiness depends on a strong data foundation and skilled talent

Article

AI can deliver only as much value as the data foundation and skilled talent supporting it. We spoke with leaders in data science, talent solutions and CIO and CISO advisory at Robert Half and Protiviti about how organizations can improve data readiness for AI, build the capabilities they need and decide where to focus their efforts first.

Data is essential for any modern organization, supporting everything from financial forecasting and operational efficiency to customer insights, workforce planning, compliance and risk management. Rapidly expanding AI use is deepening that dependence. It’s also raising the stakes for businesses to prioritize data readiness for AI by improving the accuracy, reliability and management of the information their AI tools access.

According to Robert Half’s 2026 Tech Priorities and Opportunities report, AI integration and data science, analytics, and governance rank among technology leaders’ top priorities for this year, behind only security. Additional findings underscore how AI use is accelerating: Most companies’ AI initiatives have already moved beyond the pilot stage, and 87% of small and midsize businesses (SMBs) and 97% of large enterprises are increasing AI spending this year.

As organizations embed AI in more systems and workflows, building a strong data foundation becomes even more essential. Many businesses already rely on ERP, CRM, human capital management, analytics and other platforms that increasingly include AI capabilities. But getting the most value from those tools requires integrating the data across them and improving its quality, consistency and accessibility. It also demands that technology leaders help secure and develop skilled talent, establish clear ownership for data quality and governance, and coordinate work across the organization.

In this article, leaders at Robert Half and global consulting firm Protiviti* share perspectives from their work in data science, talent solutions, and CIO and CISO advisory on how tech leaders and their organizations can build the technical, governance and workforce capabilities needed to achieve data readiness for AI.

What does a strong data foundation for AI require? A strong data foundation for AI combines accurate, accessible and well-governed information with architecture that connects systems, clear business ownership and skilled professionals who can manage the technology responsibly. Organizations can make progress by starting with a defined business use case rather than attempting to fix every data issue at once.

Why does AI need connected, well-governed data?

AI tools depend on information that is reliable, appropriately governed and accessible across the systems where business context resides. Without that foundation, organizations may struggle to scale AI beyond isolated applications. The tools may retrieve incomplete or inconsistent information, lack the context needed to produce useful results or create risks related to inappropriate data access.

“A strong data foundation gives people and AI systems trusted, connected information to make better business decisions,” says Danti Chen, senior vice president of applications, technology and innovation and head of data science at Robert Half. “When data is well governed, people know where to find it, can trust what they’re using and understand who is responsible for it.”

Those same qualities become even more important as organizations work toward data readiness for AI. “Before you put AI on top of data, you need to understand where it originated, how it has been managed, who can access it, how it changes and how all of that is governed,” says Kim Bozzella, managing director and global leader of CIO and CISO C-suite solutions at Protiviti.

The introduction of AI agents adds another layer of complexity. As agents become more sophisticated and retrieve information at greater speed and scale—and act on behalf of users or use their own identities and permissions—organizations need to know which agents are operating, what they can access and how they interact with other systems and agents.

Maintaining that visibility can be especially difficult when data is fragmented. Information may be spread across finance, CRM, operations, HR and other systems, with definitions, formats and ownership varying across the business. An AI-enabled financial platform may work well within its own environment but provide only a partial view when a business question also depends on customer, operational or workforce data stored elsewhere.

Addressing fragmentation and preparing data for AI often requires improvements to data architecture. Robert Half research finds that 39% of tech leaders at SMBs and 50% at large enterprises are prioritizing data architecture improvements in 2026. The integration of AI-driven automation and agentic and generative AI is also a priority for 53% of both SMBs and large enterprises.

These efforts are closely linked: AI is challenging to scale when the underlying data is inconsistent or hard to control. Better architecture can also reduce manual reconciliation, support consistent measures and give authorized employees faster access to the information they need.

Learn why a tech modernization strategy is critical for driving AI adoption.

Building a strong data foundation for AI: What to know to get started

Start with a business priority, not the technology
What are the biggest barriers to data readiness for AI?
What skills and roles are needed to support an AI data strategy?
Prioritize what matters most to achieve AI readiness goals

Start with a business priority, not the technology

Building an AI data strategy and a robust foundation to support it doesn’t require resolving every issue from the outset. A more practical approach is to start with a business question, decision or workflow that better data could improve—such as strengthening a financial forecast, creating a more complete customer view, streamlining an operational process or supporting an AI-enabled service request.

A focused use case helps reveal what information is required, where it resides, whether it can be trusted and who is responsible for it. “When business leaders understand the technical constraints and technology leaders understand the business goals, that’s when things go well,” Chen says. “Start with a clear use case and have a constructive dialogue about what the business is trying to achieve.”

That dialogue also helps determine who owns the data and who is accountable for the outcomes. Business leaders can set priorities and take responsibility for data quality, definitions and business context within their areas, while tech teams provide the architecture, access and controls needed to use data effectively.

Each executive’s contribution to the AI data strategy varies by role. For example:

The chief financial officer (CFO) can set financial measures and forecasting requirements.

The chief operating officer (COO) can outline key operational processes and performance indicators.

Sales and marketing leaders can surface relevant customer data and prioritize growth use cases.

The chief human resources officer (CHRO) can establish workforce data requirements and help address skills and adoption.

Security, risk and legal leaders can define access, privacy and control requirements.

The chief information officer (CIO) and other tech leaders can work with these executives to translate business requirements into decisions about architecture, integration, platforms and delivery. The chief information security officer (CISO) can help establish security, access and risk controls needed to protect the data and systems involved. But technology leaders shouldn’t be expected to take the lead alone on defining business terms, redesigning processes or determining how AI should change the organization.

“AI initiatives can fail when tech leaders say, ‘Here are some prompts, go use them,’ instead of partnering with business leaders to define the strategy first,” says Kathy Northamer, senior vice president, managed technology and digital solutions, at Robert Half. “In short, problems arise when AI is put solely on the technology team without including the business side.”

The mix of leaders involved will depend on the specific AI use case and how it affects the business. “The likelihood of achieving the intended return is much greater when the C-suite is involved from the start,” Bozzella adds. “The CIO and CISO are critical to the process, but the CHRO also needs to be involved because AI and data initiatives affect skills, roles, people and processes.”

What are the biggest barriers to data readiness for AI?

Data and AI initiatives rarely stall because of a single issue. Robert Half research shows technology leaders are managing a host of interconnected workforce and technical hurdles.

Top barriers to data and AI project execution

Team management challenges

Hiring for data science and AI skills: 63%

Balancing talent costs with AI spend: 56%

Collaborating with business units: 48%

Developing data and AI leaders: 43%

Technical challenges

Data quality and integration: 56%

Legacy systems and technical debt: 46%

Scaling AI models: 41%

Responsible AI use, compliance and privacy: 41%

These challenges are often closely connected. An organization may invest in AI before its data is ready, modernize its architecture without first resolving ownership or build analytics on inconsistent definitions and disconnected sources. In each case, the technology may be sound, but the conditions needed to use it effectively aren’t yet in place.

This risk is particularly relevant for businesses that primarily buy rather than build their technology. They may not need teams dedicated to developing proprietary platforms or AI models, but they still need professionals who can evaluate vendor capabilities, integrate and configure purchased systems, assess the quality and suitability of the data they use, and establish appropriate controls.

What skills and roles are needed to support an AI data strategy?

Executing an AI data strategy requires organizations to expand the capabilities available on their technology teams—and across the business. Along with data scientists, successful initiatives may depend on data engineers, data architects, machine learning engineers, MLOps specialists, governance leaders and professionals who can integrate AI into applications and workflows.

In a Robert Half survey, tech leaders identified these capabilities as among the hardest to secure for data and AI projects:

AI agent development and orchestration: 51%

Generative AI, including prompt engineering and LLM applications: 46%

Machine learning engineering, including productionizing models: 46%

MLOps and AI infrastructure: 45%

AI application development and integration: 45%

Data engineering, including data pipelines and platforms: 43%

“The skills gap is really around hybrid talent—people who can bridge business context, technology capability and an understanding of data,” Bozzella says. “AI doesn’t have the judgment, institutional knowledge or business context that people bring. Knowing the technology matters but understanding why the business does what it does is just as important.”

To attract and secure this hard-to-find talent, organizations may need to evolve how they hire for data and AI roles. “Many job descriptions need to be completely rewritten for today’s hiring needs around data, AI and technology,” says Northamer.

A skills-based hiring approach can help employers look beyond conventional job titles and overly narrow experience or credential requirements to assess what candidates can do, how their capabilities complement the team and how readily they can learn.

Management skills are also becoming more important as technology professionals direct, review and govern more work involving AI. “In the future, technology leaders will likely manage teams of people, teams of AI agents or a combination of both,” Chen says. “They will need to know how to manage those teams effectively, which makes people management and soft skills extremely important.”

Developing current employees is also essential to building the capabilities needed to maintain a strong data foundation and get more value from AI. Robert Half research shows many technology leaders recognize this and are already:

Providing approved AI tools for day-to-day development or operations: 76%

Enabling hands-on use of AI tools, like coding agents, copilots and AI assistants: 71%

Providing formal training for AI or automation skills: 66%

Chen, Bozzella and Northamer agree that training should be paired with opportunities to apply new skills to relevant business problems. Approved tools, defined use cases and practical guidelines can help employees gain experience while managing security, privacy and compliance risks. Organizations should also encourage teams to share effective workflows and lessons that may benefit others.

Learn more about the tech skills shortage—and 4 ways to address it head-on.

Prioritize what matters most to achieve AI readiness goals

A strong data foundation for AI doesn’t require addressing every data source and system at once, completing a full IT overhaul or building every capability in-house. Organizations can focus first on the data, technology and expertise needed for priority business use cases. This approach connects AI investments to business goals, establishes ownership for data quality and governance, and helps lay the groundwork for improving the data foundation over time.

Leaders can also decide which expertise to keep in-house and where outside support can address specific business needs or skills gaps. Permanent teams may support ongoing needs such as data governance, architecture and leadership, while contract talent and project professionals and consultants can provide additional capacity or specialized expertise for migrations, integrations, governance and other defined projects.

“Bringing in one person, or even a few people, to help get the data in order can allow organizations to start gaining value from the AI capabilities in technology they probably already own,” Bozzella says.

With the right technology, governance structures and talent in place, organizations can strengthen their AI readiness while improving decision making, operations, customer experiences and compliance.

Go to Tech Insights

Explore Robert Half’s Tech Insights page for more research on technology skills gaps and workforce strategies. You can also download the 2026 Tech Priorities and Opportunities report for guidance on building tech teams that can deliver on strategic priorities.

*Protiviti is a Robert Half subsidiary.

Technology modernization strategy: How to build the right team in 2026

11 Key Players Powering Digital Projects—And the AI Tools Behind Their Success

AI, Automation and Upskilling: How Tech Leaders and Professionals Can Stay Ahead