State of AI in Procurement in 2026

Procurement leaders have spent the last two years being told that AI will take the manual work off their plates. Fresh research says most teams are not there yet in 2026. In a new survey of 73 procurement executives by HFS Research in partnership with Art of Procurement, only 15 percent say they have fundamentally reimagined procurement with AI at the core, while 38 percent are still running only pilots and proofs of concept.

That gap between ambition and day-to-day reality is the clearest signal in this year’s update of our State of AI in Procurement, and it runs through the adoption data, the use cases, the barriers, and leaders’ expectations for the next three to five years.

Key findings: AI in procurement statistics for 2026

Our 2026 data comes from two kinds of sources. The first is new, first-party research. HFS Research, in partnership with Art of Procurement, surveyed 73 procurement executives across nine industries. Ninety percent of respondents are VP-level or above, and 78 percent work for companies with $1 billion or more in annual revenue. HFS Research president Saurabh Gupta presented the interim results at Art of Procurement’s Catalyst | San Francisco on May 5, 2026. The second is other 2026 research from The Hackett Group, ProcureCon, Gartner, and others, with each source linked.

New research from Art of Procurement and HFS Research

• 15 percent of procurement organizations have fundamentally reimagined procurement with AI at the core. Source: HFS Research in partnership with Art of Procurement, 2026 (interim data, 73 executives), presented at Catalyst | San Francisco.
• 38 percent are still running only AI pilots and proofs of concept in procurement. Source: HFS Research in partnership with Art of Procurement, 2026.
• 95 percent agree AI is reshaping roles faster than procurement teams are redesigning their operating model. Source: HFS Research in partnership with Art of Procurement, 2026.

More fresh AI in procurement statistics and research

• 43 percent of organizations are actively pursuing AI deployment in procurement, nearly double the level reported a year earlier. Source: 2026 Procurement Key Issues Study, The Hackett Group.
• 36 percent of CPOs are very confident in their ability to redesign roles and processes around AI. Source: Gartner, May 2026.
• 80 percent of procurement executives identify AI-enabled technology as the most transformational trend for the function over the next five years. Source: 2026 Procurement Key Issues Study, The Hackett Group.
• 69 percent of organizations access AI through capabilities embedded in their existing procurement platforms, particularly for transactional processes. Source: 2026 Procurement Key Issues Study, The Hackett Group.
• 65 percent of procurement leaders are piloting AI and 24 percent are still evaluating AI opportunities. Source: 2026 ProcureCon CPO Report.
• IT and Purchasing/Procurement lead all business functions on both frequency of generative AI use and confidence; 82 percent of enterprise leaders overall use generative AI at least weekly. Source: Accountable Acceleration, Wharton and GBK Collective, 2025.
• About 900 US searches a month for “ai in procurement” in the year to September 2026, up from about 145 four years earlier; searches for “agentic procurement” rose from zero in early 2025 to more than 120 a month. Source: Art of Procurement analysis of Ahrefs Keywords Explorer data, US.
• 26 percent name contract creation and analysis as the area where AI is most actively used, followed by category strategy and intelligence (18 percent) and spend analytics (18 percent). Source: HFS Research in partnership with Art of Procurement, 2026.
• 49 percent say the most exciting benefit of AI is a better employee experience (26 percent) or a more strategic focus (23 percent); cost savings and efficiency ranks third at 22 percent. Source: HFS Research in partnership with Art of Procurement, 2026.
• 50 to 80 percent of current procurement work could be automated by generative AI, based on KPMG simulations. Source: The Future of Procurement: Gen AI’s Impact, KPMG.
• 25 to 40 percent is how much agentic AI could increase procurement efficiency. Source: Transforming Procurement Functions for an AI-Driven World, McKinsey.
• 15 to 30 percent efficiency improvements are already being captured by procurement teams using AI category agents to automate non-value-added activities. Source: The Future of Category Management, McKinsey.
• 8 percent is the expected increase in procurement workloads in 2026, even as head count and operating budgets decline. Source: 2026 Procurement Key Issues Study, The Hackett Group.
• 18 percent name legacy technology as the biggest challenge in adopting AI, ahead of poor data quality (16 percent) and AI technology not being ready (15 percent). Source: HFS Research in partnership with Art of Procurement, 2026.
• 22 percent say lagging governance is the top unintended risk emerging from AI in procurement. Source: HFS Research in partnership with Art of Procurement, 2026.
• 67 percent of CPOs cite data privacy, security, and compliance as a barrier to scaling AI. Source: 2026 ProcureCon CPO Report.
• 57 percent of CPOs rank siloed ways of working among the top barriers to value delivery, the most cited barrier. Source: 2025 Global CPO Survey, Deloitte.
• 96 percent see the future of procurement as human plus machine; 62 percent expect a human-led model with AI support and 34 percent expect AI to lead with human oversight. Source: HFS Research in partnership with Art of Procurement, 2026.

What exactly is AI in procurement?

AI in procurement is the use of artificial intelligence to automate and support procurement work, from spend analysis and sourcing to contract management and supplier risk monitoring. It shows up in two ways: AI features embedded in procurement software, and general-purpose generative AI tools that procurement professionals use for everyday tasks.

Both routes are now common. The Hackett Group’s 2026 research found that 69 percent of organizations access AI through capabilities embedded in their existing procurement platforms, particularly for transactional processes. Individual use is high as well. In the 2025 edition of the Wharton and GBK Collective adoption study, IT and Purchasing/Procurement led all business functions on both frequency of generative AI use and confidence, at a time when 82 percent of enterprise leaders overall used generative AI at least weekly.

Interest from the wider market has grown in step. Art of Procurement’s analysis of Ahrefs US search data shows average monthly searches for “ai in procurement” rising from about 145 in the year to September 2022 to more than 1,100 in the year to September 2025. Searches for the head term have since leveled off at roughly 900 a month, while searches for “agentic procurement” climbed from zero in early 2025 to more than 120 a month by autumn 2026. The conversation is moving from whether to use AI to how much of the work AI agents should carry.

Types of AI used in procurement

Artificial intelligence covers several technologies, and procurement teams meet most of them:

• Machine learning (ML) learns from data without being explicitly programmed. It powers spend classification and anomaly detection.
• Natural language processing (NLP) lets software read and generate human language, which is how contract and supplier documents become usable data.
• Generative AI produces new text and analysis. It is a specialized application of deep learning, which processes information in layers to find patterns in unstructured data.
• Agentic AI describes systems that can perceive their environment, make decisions, and take actions toward a goal with limited human input.

Robotic process automation (RPA) often appears in the same conversations. Rule-based RPA is generally not considered AI, although many platforms now combine the two.

How AI is used in procurement organizations today

AI adoption in procurement is broad but shallow. Most teams use AI somewhere, and few have redesigned how work gets done around it.

Most procurement teams are still in pilot mode

When HFS Research and Art of Procurement asked how AI is being deployed in the procurement function, the largest group (38 percent) said they run only pilots and proofs of concept.

Other 2026 research points the same way. The Hackett Group found that only 12 percent report large-scale implementation, with most organizations still operating pilots or single-use-case deployments. The 2026 ProcureCon CPO Report found that 11 percent are implementing AI with measurable impact, 65 percent are piloting, and 24 percent are still evaluating opportunities. HFS Research summed up the pattern in its Catalyst | San Francisco presentation as “death by a 1,000 POCs.”

The top AI use cases for procurement are strategic

The same survey asked where AI is most actively used across the procurement value chain. Contract creation and analysis leads at 26 percent, followed by category strategy and intelligence (18 percent), spend analytics and insights (18 percent), purchasing and transactions (12 percent), risk monitoring and compliance (10 percent), supplier performance management (8 percent), and sourcing and supplier selection (6 percent).

Together, contracts, category strategy, and spend analytics account for 62 percent of active use. In practice, teams are using AI in two ways:

• Ad-hoc generative AI tasks, such as drafting documents and emails, developing scopes of work, answering RFP questions, analyzing RFP responses to support scoring, building negotiation plans, and drafting contract clauses. Our article on ChatGPT for procurement covers these in more detail.
• AI built into procurement workflows and tools, which covers the applications below.

Seven common AI applications in procurement

1. Spend analytics

Machine learning was one of the first AI applications in procurement, starting with spend classification. Today’s AI-powered spend analytics classify spend automatically, surface savings opportunities across business units, flag maverick spend and compliance issues, and combine internal and external data for near real-time visibility.

2. Semi-automated sourcing

AI can help stakeholders run parts of the sourcing process themselves. In AI-enabled sourcing and contracting solutions AI turns informal requirements into structured scope documents, analyzes and scores RFP responses, and drafts contract language, SLAs, and KPIs.

3. Procurement orchestration and intake

Intake tools and orchestration software act as the front door for employees who want to buy something. A user types what they need, AI matches the request to categories and backend systems, and the tool guides the user through the right approval path.

4. Contract management

AI extracts key terms and obligations, identifies risk in contract language, monitors compliance, and supports renewals. In contract management tools, this turns contract metadata into information procurement can act on.

5. AI-based negotiation

Providers such as Pactum run automated negotiations with cohorts of suppliers, covering payment terms and other contract terms as well as price. This can bring negotiation to tail suppliers that procurement teams rarely have time to engage.

6. Supplier management

Across supplier management, AI improves supplier data quality and supports risk assessment from multiple data sources. It also monitors performance and identifies new suppliers or consolidation opportunities.

7. Strategic decision support

AI combines internal and external information to inform sourcing strategy and supplier selection. Teams get market intelligence they can act on before a decision is made.

Where AI procurement software fits

Most of these applications reach procurement teams through software they already own or are evaluating. To compare the solution providers behind them, see the Procurement Provider Directory and our list of procurement software providers.

Benefits and value of AI in procurement

What procurement leaders value most

Cost savings ranks third among the benefits procurement leaders find most exciting. Asked to name the most exciting benefit of AI, respondents to the HFS Research and Art of Procurement survey chose a better employee experience (26 percent), strategic focus (23 percent), cost savings and efficiency (22 percent), faster decision-making (12 percent), better supplier insights (7 percent), and improved risk visibility (6 percent). Four percent said it is too early to say.

Nearly half (49 percent) point to productivity-related outcomes ahead of cost. That matches a second finding from the survey: only one in four procurement leaders say they have a consistent and influential seat at the enterprise strategy table. HFS frames procurement as “not excluded from strategy,” but “trapped in execution,” buried under transactions, processes, and firefighting. AI is the most realistic way to give that time back.

1. Efficiency gains

Automating routine tasks shortens procurement cycles and frees people for higher-value work. The Hackett Group expects procurement workloads to increase by 8 percent in 2026 even as head count and operating budgets decline, which makes this capacity more than a nice-to-have.

2. Cost reduction

AI augments the analytical capability of procurement professionals by identifying savings that people miss and supporting category management with data. Automation also lowers processing costs and errors.

3. Risk mitigation

AI can monitor supplier financial health alongside geopolitical and market signals. The result is earlier warning of supply disruption or fraud.

4. Fact-based supplier relationships

With better data, procurement teams can communicate more consistently with suppliers and base development plans on performance. Shared information also gives both sides a firmer footing for partnership.

Challenges and barriers to AI adoption in procurement

Asked for the single biggest challenge in adopting AI, survey respondents named legacy technology (18 percent), poor data quality (16 percent), AI technology not being ready (15 percent), lack of governance (14 percent), resistance to change (11 percent), limited trust in AI outputs (10 percent), operating model misalignment (6 percent), and broken processes (4 percent). HFS advises leaders to “stop obsessing over AI tech” and first fix everything around it that is broken.

Data quality issues

Inconsistent formats, incomplete history, weak taxonomies, and poor master data all limit what AI can do. Data will never be 100 percent accurate, so procurement teams should mitigate the issues that matter for their priority use cases and get started.

Legacy systems and integration

Multiple ERP instances and disconnected procurement tools make it hard for AI to see the full picture. AI cannot solve integration problems on its own, and it needs secure access to the data it is given.

Security and compliance

Data privacy, security, and compliance top the list of barriers to scaling AI in the 2026 ProcureCon CPO Report, cited by 67 percent of respondents. Many IT departments now direct employees to licensed enterprise tools over free consumer versions.

Validating AI outputs

AI outputs still need checking. Procurement teams need a trust-but-verify approach and the judgment to know when an output should be questioned.

Change resistance and skills gaps

Fear of job displacement and comfort with existing tools both slow adoption. Skills are a related gap: BCG found that only 6 percent of senior leaders have begun upskilling their workforce in a meaningful way.

Best practices for AI adoption in procurement

Start with problems, not tools

Chris Sawchuk, principal and global Procurement Applied Intelligence practice leader at The Hackett Group, puts it plainly:

“Don’t lose sight of why you’re implementing or investing in a particular area. Everyone has to answer the questions of how you’re integrating AI into your product; but the question is, how is it impacting the things that you were doing before, and how is it making that even better in this new environment?”

Anchor each AI initiative in a measurable outcome, such as cycle time, spend visibility, contract compliance, or supplier risk. Define that outcome before the pilot starts, so the team knows what success looks like when it is time to scale.

Build the business case beyond cost savings

HFS Research recommends evaluating AI initiatives against four measures: productivity, prediction, personalization, and performance. Across 979 GenAI and agentic AI use cases HFS collected over 12 months, productivity accounted for 54 percent of pilots but only 19 percent of use cases in production, while performance outcomes such as growth and competitive advantage rose from 8 percent of pilots to 27 percent of production use cases. Initiatives justified on efficiency alone are the ones that tend to stall.

Close the gap between governance and adoption

In the HFS Research and Art of Procurement survey, 95 percent of procurement executives agree that AI is reshaping roles faster than they are redesigning their operating model. Asked which unintended risks are emerging, they named governance lagging (22 percent), data quality distorting outcomes (12 percent), procurement’s influence not increasing (11 percent), over-reliance on AI (11 percent), efficiency gains that are not strategic (10 percent), and decisions that are harder to audit (10 percent). Gartner’s research shows the same concern from another angle: only 36 percent of CPOs are very confident they can redesign roles and processes around AI.

Governance works best when it is cross-functional from the start. Deloitte’s 2025 Global CPO Survey identifies siloed ways of working as the most-cited barrier to value delivery, cited by 57 percent of CPOs. An AI governance committee with procurement, IT and data, legal and compliance, risk, and finance, sponsored jointly by the CPO and CIO, keeps technology roadmaps and accountability aligned. My AI Governance Framework for Procurement covers accountability, transparency, fairness, risk management, and data governance in more detail.

Fix the data your priority use cases need

Teams do not need perfect data to start. Supplier master data cleanup is a practical first step, and spend categorization across business units usually comes next. Focus on the data domains your priority use cases depend on.

Invest in your people’s AI skills

AI adoption is a people challenge as much as a technology challenge. Capability building starts with shared AI literacy across the team and the adoption of tools in daily workflows. Over time, it extends to redesigning processes around what AI can do. HFS calls the gap between how quickly individuals adopt AI on their own and how slowly enterprises move the “AI velocity gap.”

Lessons from MIT: avoiding common AI pilot traps

MIT’s research on the state of AI in business, reported by Forbes, found that 95 percent of corporate AI initiatives show zero return. Employees at more than 90 percent of surveyed companies already use personal AI tools, and AI projects built with external partners reach deployment about twice as often as internal builds (around 67 percent compared with roughly 33 percent).

For procurement, the first lesson is to anchor projects in measurable outcomes and to formalize and secure the “shadow AI” tools teams already use. It also pays to favor systems that learn from procurement workflows and to treat solution providers as partners who share accountability for results.

The future of AI in procurement

Procurement leaders expect people to stay in charge. Asked about the role of AI in procurement over the next three to five years, 62 percent of HFS Research and Art of Procurement survey respondents expect a human-led model with AI support, 34 percent expect AI to lead with human oversight, and only 3 percent expect procurement to be highly autonomous. In total, 96 percent see the future as human plus machine.

AI agents and virtual procurement advisors

AI agents act as virtual team members that draw on spend history and market conditions. Category management is a likely early use. McKinsey reports that procurement teams using AI category agents are already capturing 15 to 30 percent efficiency improvements by automating non-value-added activities, and that agentic AI could increase procurement efficiency by 25 to 40 percent.

Autonomous sourcing for routine categories

Systems are moving toward running full sourcing cycles for routine categories, from spotting an opportunity in upcoming contract expirations through to monitoring supplier performance after award. Strategic suppliers and complex categories will stay with people.

Supplier and market intelligence

AI is making supplier and market intelligence more personal, with alerts tailored to a category portfolio and predictive supplier performance analytics. Scenario modeling can also translate geopolitical events into procurement risk.

Embedded sustainability

Automated ESG scoring, carbon calculations at the sourcing stage, and regulatory monitoring are bringing sustainability into everyday sourcing decisions. That moves it beyond the annual reporting cycle.

The bottom line on the state of AI in procurement

AI is now part of how procurement works, but the 2026 data shows a function caught between ambition and reality. Most teams use AI somewhere, few have redesigned work around it, and governance has not kept pace with adoption. The organizations that break out will treat AI as an operating model question. They will pick outcomes before tools, build business cases that go beyond savings, fix the data their priority use cases need, and invest in their people as seriously as their platforms. Procurement leaders expect humans to stay in the lead, so the job now is to prepare those humans to work alongside AI.