AI and the future of procurement and contracting

As agencies across the government focus on strategies for enhancing efficiency and productivity, as well as reducing waste, automation and AI are key drivers in transforming how the government carries out its missions. While automation is already prevalent in many day-to-day operations, government leaders are shifting focus to procurement and contracting, exploring ways to make contracting more efficient. 

A recent executive order aimed at consolidating procurement underscores the current administration’s commitment to streamlining government contracting. The General Services Administration (GSA) is leading this evolution with the goal of reducing redundancy and waste between agencies. Google quickly embraced these efforts by working with the GSA to offer Google Workspace and an advanced Gemini for Government AI platform to deliver industry-leading cloud, AI models, agentic solutions and more to every federal agency at a reduced price.

“GSA is taking another look at how distributed procurement has been,” said Chris Niehaus, strategic growth executive at Google Public Sector. “The government is saying, ‘We are a big entity, we should buy in a uniform way, rather than a fragmented way.’ That creates a lot of inefficiency, varying pricing and a lot of contracts to manage.”

Standardizing and streamlining procurement

Consolidating procurement, however, is only the first piece of the puzzle. The federal government needs ways to speed the overall contracting pipeline for the GAO workforce, now managing a much larger workload on behalf of the entire government. 

“There is a lot of work going on to standardize procurement, particularly around two big efforts: FAR 2.0 and FedRAMP 2.0x,” Niehaus said.

The former, Federal Acquisition Regulation (FAR) 2.0, is intended to “return the FAR to its statutory roots, rewritten in plan language, and remove most non-statutory rules,” to enable faster acquisitions, greater competition and better results. Meanwhile, FedRAMP 2.0x seeks to give agencies faster access to new technologies by speeding the FedRAMP accreditation process. 

Ultimately, all of these efforts represent the government working toward achieving the same rate and pace of innovation as the commercial sector — without sacrificing security and compliance.

“That’s really matching the way Google has built out its cloud. We built our government cloud in our commercial cloud instead of building a secondary government cloud,” Niehaus said, “so we can bring AI and other capabilities to our government customers at the same time as our commercial with a software-defined set of compliance tools.”

Use cases for AI in the contracting lifecycle

Of course, updated policies and statutes can only get the federal workforce so far without technology tools to assist in making daily operations more efficient, particularly in the current environment of pared down staffing for many agencies. This is where AI and automation take center stage.

Current use cases for government contracting include:

• AI for opportunity identification and market research: Scanning databases and sites for contract opportunities while analyzing data, pricing and competitor trends tailored to the strengths of the provider.
• AI assistance in proposal development: Generating proposal outlines, drafting standard sections, ensuring compliance with requirements and tailoring content to vendor’s past performance. 
• AI tools for compliance automation: Scanning proposals and contract documents for adherence to regulations like FAR/DFARS, flagging missing elements, scoring, and recommending remediations. 
• AI for contract management and analysis: Monitor contract performance, tracking deadlines, deliverables and spending patterns for efficiency. 

While these areas undergo AI transformation, there are even more advanced, sophisticated use cases coming down the pike. Agentic AI could be used to create AI agents engineered for multi-stage processes, from proposal development and compliance check through approval reviews. 

AI infrastructure has the potential to expand beyond small use cases or single models to support many models with the security, tools, manageability and resources for scaled development and deployment. 

Finally, multimodal AI — not just text but code, video and audio — could be integrated into AI assistants and agents to evaluate the quality of code, for example, and score prototypes rather than just written proposals. 

Speaking the same language

On top of identifying the best use cases, an open-source, standards-based approach to automation in the contracting process is critical to efficiency and effectiveness. To that end, NIST, FedRAMP and industry leaders established the Open Security Controls Assessment Language, or OSCAL, a machine-readable language that leverages automation to standardize security assessments. 

As part of the technical advisory group for the OSCAL Foundation, Google is playing an integral role in driving the organization’s mission forward. In 2023, Google announced it had adopted OSCAL taxonomy internally to ensure Google Cloud security controls are consistently assessed and make it easier to automate security posture assessment.  

“[OSCAL] really simplifies standards around security assessment and exchange of information, and I think we’re going to see more like it show up in FAR 2.0 and FeRAMP 2.0X to increase automation and accelerate AI,” Niehaus said.

Learn more about how Google Public Sector is working to increase efficiency and productivity across the government procurement and contracting pipeline.