As Artificial Intelligence (AI) increasingly permeates various sectors, its integration into the intricate landscape of federal contracting presents both transformative opportunities and significant challenges. While contractors are rapidly adopting AI-powered tools to streamline proposal development and submission, the government procurement evaluation systems often remain anchored in manual, human-centric processes. This growing disparity, termed the “automation mismatch,” poses critical questions about efficiency, fairness, and the future of public sector procurement.
Exploring this complex intersection is Farhan Amjad, a Technical Analyst at Intellect Solutions LLC. Known for his work leading enterprise automation, AI integration, and federal digital modernisation initiatives, Amjad brings a unique interdisciplinary perspective combining rigorous software engineering and project management with deep public-sector insight. His expertise spans cloud technology, compliance, and digital transformation, positioning him as a trusted voice at the frontier of this evolving field.
Amjad’s recent research, notably explored in his work “Automation Mismatch: How Contractor AI Adoption Challenges Institutional Procurement Norms,” dives directly into how federal procurement systems struggle to keep pace with the accelerating adoption of AI by vendors. He argues that this creates structural tension that demands careful examination and strategic intervention.
AI’s Pervasive Presence in Amjad’s Work
For Amjad, AI is no longer a standalone technology but an embedded component across his work at Intellect Solutions. His role involves engaging with AI on multiple levels, from hands-on implementation to critical research and evaluation.
“AI is no longer confined to a single toolset—it’s embedded across everything we do, from automation to analytics to content generation,” Amjad explains. “In my role, I engage with AI on multiple levels. I work hands-on with automation platforms like UiPath for intelligent workflow automation, and I’ve prototyped AI-powered triage engines using open-source tools like H2O.ai to explore how procurement data can inform strategic decisions.”
His work also involves supporting experimentation with scalable AI deployments within AWS cloud environments, ensuring alignment with stringent federal security standards. Beyond the technical implementation, Amjad approaches AI from a researcher’s perspective, examining how generative AI and machine learning can be responsibly integrated into public sector processes, particularly in areas where regulation, transparency, and fairness are paramount.
Amjad points to diverse applications already in play:
“Whether it’s RPA bots handling onboarding, language models drafting intake reports, or predictive models flagging high-value solicitations, AI is everywhere. But that makes the question of how we use it even more critical. My job is to help design, test, and evaluate AI implementations through both a technical and ethical lens—ensuring they’re not only efficient, but also institutionally appropriate and publicly accountable.”
Amjad’s practical experience underscores this point. He is the lead architect behind Intellect’s Vendor Intelligence Database and serves as the firm’s Salesforce Administrator and RPA Program Lead—initiatives that have directly contributed to contract wins and enterprise-wide efficiencies, demonstrating the tangible impact of AI and automation in the contractor sphere.
The Current State: An Asymmetry in Federal Contracting
According to Amjad, the current landscape of AI in federal contracting is characterised by a significant asymmetry between contractors and government agencies. “We’re at a moment of asymmetry,” he said.
On the contractor side, AI tools are actively reshaping operations. Solutions like Vultron and Unanet ProposalAI are widely used to accelerate and refine proposal development, enabling vendors to pursue and submit more bids with increased speed and often improved compliance accuracy. This allows contractors to scale their responses and potentially gain a competitive edge.
However, the government side presents a stark contrast. Agencies, even those with forward-leaning technology mandates, often remain constrained by established, human-centred, and regulation-bound evaluation systems. Amjad’s research highlights that despite encouraging policy guidance from entities like the Office of Management and Budget (OMB) and the White House, most federal evaluators face substantial cultural, regulatory, and operational hurdles in adopting AI, particularly within the procurement function.
Amjad points to specific legal frameworks, such as Federal Acquisition Regulation (FAR) 15.303 and 15.304, which mandate documented human justification for evaluation criteria and source selection decisions. These requirements make the integration of many machine learning models—especially those employing dynamic feature weighting or complex, opaque algorithms—challenging to deploy within the existing legal and procedural framework.
“So, while contractors are accelerating with AI, agencies are still trying to reconcile innovation with fairness, transparency, and oversight,” Amjad summarizes. “That’s the state of the field: technically advanced, institutionally hesitant.”
Achieving Efficiency: Conditional and Constrained
While the broader adoption of AI in government procurement evaluation lags, Amjad acknowledges that AI is already contributing to efficiency in certain pockets of government operations. These applications are often focused on reducing administrative burdens through automation.
“In pockets, AI is already helping government agencies reduce administrative burden and improve speed—particularly through Robotic Process Automation (RPA) and document processing,” Amjad notes. He cites examples like the IRS and GSA leveraging AI to streamline processes such as responding to Freedom of Information Act (FOIA) requests, auditing logs, and reviewing procurement files.
However, these existing uses are typically confined to specific, low-risk scenarios. “These are typically constrained use cases—where the risk of bias or lack of transparency is low,” Amjad explains. He emphasises a critical point: “AI doesn’t inherently make government more efficient. It makes it conditionally more efficient—when embedded in processes designed for traceability, fairness, and legal defensibility.”
This conditional efficiency explains why much of the current AI adoption in government remains narrow, often driven by specific cost-saving imperatives rather than broader mission innovation or transformative process redesign. Though efficiency gains are real in these specific applications, they are scattered and do not yet represent a systemic shift in core functions like complex procurement evaluation. According to Amjad’s research, achieving sustainable efficiency requires concurrent institutional change alongside technical deployment. Without this parallel evolution, AI risks remaining a ‘bolt-on’ feature rather than a true transformative force within government.
Identifying and Bridging the “Automation Mismatch” Gap
Amjad identifies the “automation mismatch” as the most urgent gap requiring attention. This structural divide between the AI-enhanced capabilities of federal contractors and the predominantly manual evaluation systems used by government agencies creates significant operational strain on both sides and can potentially undermine the fairness and effectiveness of the procurement process.
“Contractors use AI to submit more proposals at higher speed, while contracting officers are still expected to narratively justify every decision using rigid, human-readable criteria,” Amjad elaborates. This imbalance directly impacts evaluation timelines, contributes to backlogs, and could inadvertently disadvantage innovative approaches that don’t easily fit into traditional, human-evaluated frameworks.
Bridging this gap, Amjad argues, necessitates more than simply deploying better technology within government agencies. It demands fundamental institutional alignment. His research underscores that existing regulatory frameworks, such as the FAR and agency-specific manuals like the Health and Human Services Acquisition Manual (HSAM), must evolve in parallel with advancements in AI tooling.
Beyond regulatory reform, addressing the gap requires practical steps like providing adequate training for government evaluators on understanding and potentially utilising AI insights, developing clear policy guidance on the role and acceptable use of explainable AI (XAI) in evaluation, and fostering robust cross-functional collaboration between IT specialists, legal counsel, and procurement teams. “Without addressing these institutional conditions, the AI gap will persist—even if the technology keeps advancing,” Amjad warns.
The Future: Context-Aware, Transparent, and Integrable AI
Looking ahead, Amjad is optimistic about the potential for AI to better serve the federal government, but he grounds this optimism in the necessity of continued development focused on traits essential for public sector use: context-awareness, transparency, and integrability.
Improvements in explainable AI (XAI) models are particularly promising. These models are designed to allow human users, such as procurement officers, to understand the rationale behind an AI system’s recommendations or decisions. Tools that can map AI-driven insights back to specific solicitation criteria in a clear, human-readable format will be crucial for meeting regulatory requirements and building trust.
Amjad also anticipates that the development of low-code AI platforms will democratise access to AI capabilities within agencies, enabling a wider range of teams—not just specialised data scientists—to experiment with and implement intelligent automation solutions. As AI models become more sophisticated and regulatory bodies provide clearer, more definitive guidance on their use in government, there is potential for the development and adoption of more mission-aligned AI solutions tailored to the unique needs of specific agencies, moving beyond generic enterprise tools.
However, Amjad consistently returns to his central thesis: the real-world impact of these technological improvements will ultimately be determined by institutional readiness. He stresses that organisations must build internal capacity, adapt processes, and evolve policies in parallel with technological adoption.
In the most favourable future scenario, AI will move beyond merely supporting procurement functions to fundamentally reshaping how agencies deliver services, allocate resources, and even inform policy planning. But realising this potential requires a deliberate and concerted effort to bridge the ‘automation mismatch’ and ensure that institutional reforms keep pace with technological innovation, guided by experts like Farhan Amjad who possess the dual perspective of technical capability and public sector understanding.
