**Washington**: The Office of Management and Budget has issued a memo outlining new deadlines and guidelines for federal agencies on AI procurement, aiming to improve oversight and reduce risks associated with contract management, amid concerns about practical implementation and vendor lock-in challenges.
The Office of Management and Budget (OMB) has recently issued a memo outlining new guidelines for federal agencies regarding the acquisition of artificial intelligence (AI) capabilities. This initiative reflects the Trump administration’s acknowledgement of the evolving landscape of AI procurement across government sectors. The memo contains four key deadlines over the next year aimed at improving oversight and establishing clearer expectations after contracts are awarded.
Kraig Conrad, CEO of the National Contract Management Association, highlighted the memo’s significant focus on the post-award phase of federal contracts. “In any contractual relationship anywhere, whether it’s public procurement, federal or commercial, the post-award is where dollars are destroyed,” Conrad commented in an interview with Federal News Network. He noted that regular check-ins are essential, given that contract value deteriorates during this phase. The memo proposes that agencies should establish a reporting timeline, suggesting updates every five months or quarterly, to ensure accountability and monitor project progress effectively.
The OMB memo specified four deadlines for agencies, including two related to the General Services Administration (GSA) and two for all federal agencies. The GSA has been tasked with developing a public guide to assist agencies in AI procurement within 100 days. By 200 days, GSA must also create an online portal aimed at federal agencies, which will provide resources such as sample solicitations and standard contract clauses for commonly used AI systems.
As for the federal agencies themselves, they have two main deadlines: one at 270 days for updating internal acquisition procedures to reflect the new guidelines and another at 200 days for ensuring proper procedures and contract terms regarding government data use and intellectual property rights. The OMB has advised agencies to carefully consider these rights, especially concerning the data used to train AI systems.
Despite the clarity provided by the memo, experts have raised concerns about its practical implementation. One anonymous industry executive expressed worry that many deadlines extend into fiscal year 2026, leaving a gap for AI purchases made in the current fiscal year. This could result in inconsistency as agencies interpret and apply the guidelines. Conrad noted that while the emphasis on avoiding vendor lock-in is crucial to modern procurement practices, significant challenges remain.
John Ferry, president of Trenchant Analytics, welcomed the ongoing focus on preventing vendor lock-in in AI purchases. “At a more granular level, that speaks to experiences that government has had in the past,” he remarked. However, he noted that agencies face challenges in ensuring the accuracy of data for AI implementations, underscoring the complexities of managing intellectual property and data rights.
One notable change in the new memo is its recommendation for specific types of contracts when acquiring AI capabilities. Instead of traditional fixed-price contracts, agencies are encouraged to utilise statements of objectives and performance-based contracts with incentives. Conrad acknowledged this shift towards more agile contracting methods, which allows for iterative learning and adaptability throughout the procurement process.
Nevertheless, experts cautioned that these new approaches may present difficulties for agencies accustomed to standard contracting practices. The executive noted that incentive-based contracts can be complicated, particularly in terms of explaining their rationale to auditors, which could lead to inconsistencies in application.
To alleviate some of the burdens presented by contracting processes, OMB has advised agencies to leverage AI technology within the procurement process itself. The pilot programme led by the Department of Homeland Security showcased AI applications for market research and contract close-outs, while the GSA has begun testing generative AI tools to aid acquisition professionals in navigating policies and directives.
Ferry pointed out that integrating AI into acquisition processes is particularly feasible due to the standardised and linear nature of procurement activities. “Acquisition is something that affects everybody. It is an ideal use case for the acquisition workforce,” he explained, highlighting the pressing need for efficiency amid shrinking personnel numbers and increasing budget demands.
The guidance from the OMB marks a significant step in modernising the acquisition of AI tools within the federal government. However, it also raises questions about implementation timelines, agency capability, and the implications of shifting contractual methodologies. As the landscape of AI continues to change rapidly, the impact of these guidelines will unfold in the coming years.
Source: Noah Wire Services
Subscribe to Industry Updates
