In an interview with The American Bazaar, G-P’s Head of HR Laura Maffucci explains why human oversight remains indispensable, why global talent is becoming even more critical in the AI era, and what the shift from AI hype to practical implementation means for professionals and the future of work.
For thousands of tech workers across America, especially immigrants on H-1B and other employment visas, the last two years have been marked by uncertainty. Waves of layoffs have swept through Silicon Valley and Corporate America, while the rapid rise of artificial intelligence has fueled fears that entire categories of white-collar jobs could be automated away. For foreign-born professionals whose immigration status is often tied to their employment, the stakes are even higher: losing a job can mean losing the right to remain in the country.
Against this backdrop of anxiety, a new report suggests that the AI revolution may be entering a more pragmatic phase. Companies that rushed to embrace artificial intelligence are now demanding proof that the technology can deliver real business value. A 2026 AI at Work Report by G-P (Globalization Partners) an employment leader ranked No. 1 by industry analysts, that helps companies hire and manage talent in 180+ countries through its AI-powered workforce platform offers an important insight. The report reveals that 100 percent of surveyed executives say they are using AI, yet nearly 70 percent are prepared to reduce AI spending if it fails to meet expectations this year. The report found that 73 percent of executives believe the return on investment from AI has fallen short of expectations, signaling a shift from AI hype to what the company calls a “high-stakes reckoning.”
READ: As anti-immigrant hate rises, California’s $100 million Stop the Hate funding faces expiration (June 22, 2026)
The American Bazaar catches up with Laura Maffucci, Vice President and Head of HR, G-P. Maffucci oversees the global workforce, talent, and employee experience and is a respected professional with over 20 years of experience in HR. A staunch advocate for the employee experience and creating a culture of inclusivity, Maffucci talks about the “human-in-the-loop” model and humans leading the work
Your report finds that 77% of executives lack full confidence in AI accuracy. What’s driving this trust gap despite the rapid adoption of AI tools?
The past few years were defined by a rush to adopt AI, driven by competitive pressure and market excitement. In 2026, we are witnessing an AI reckoning as operational realities set in. G-P’s 2026 AI at Work Report found that 69% of leaders spend more time monitoring and acting as high-level editors for unvetted AI. We are moving away from unchecked experimentation. Unless technology is purpose-built for complex, highly specialized functions, and grounded in that expertise, generic AI tools simply cannot guarantee accuracy. Leaders are realizing that true operational confidence requires a “human-in-the-loop” model. By pairing purpose-built technology with human professionals who understand contextual nuance and regulatory risk, organizations ensure that speed doesn’t compromise quality.
Where do organizations currently draw the line between AI assistance and human judgement?
Organizations are actively recalibrating the balance between AI efficiencies and human oversight. In fact, we saw a 20-point drop in aggressive AI approaches this year compared to our 2025 data. Corporate leaders are no longer innovating purely for the sake of innovation. Instead, they demand proven, low-risk use cases.
The line is explicitly drawn at accountability. Our report found that 60% of executives utilize a model where humans lead the work and AI is integrated purely to provide data and speed. Even among the 17% who use AI to drive core processes, humans remain in the loop to handle exceptions and oversee the system. From an HR perspective, this shift highlights a growing maturity as executives view AI as a powerful cognitive forklift, not the driver.
How should employers measure meaningful AI usage?
This is a critical cultural challenge for leadership and HR teams. This anxiety exists because many organizations mandated or rushed AI adoption without providing clear business cases, usage guidelines or training.
READ: Maryland District 23 candidate Raaheela Ahmed on education, transparency, and public service (June 11, 2026)
To counter this, employers and HR leaders must establish a top-down strategy that clearly outlines how performance is defined and measured in an AI-assisted workplace. This requires moving away from superficial activity metrics such as the volume of emails or code generated, and shifts the focus to strategic, high-value outcomes. Because the true value of AI lies in freeing up human capacity rather than simply producing more work, organizations should evaluate how effectively employees reallocate their recovered time toward more strategic initiatives.
By implementing these clear guardrails and shifting to value-centric metrics, leadership teams create the psychological safety necessary for employees to experiment and learn. When workers understand that AI is an enterprise tool meant to elevate their roles rather than replace them, the urge to use technology as a performative productivity prop naturally disappears.
Nearly half of executives say they would not trust AI with risk management, financial decisions, talent strategy, or compliance. What are the biggest concerns in these high-stakes areas?
In these high-stakes domains, the margin for error is virtually zero. A single misstep can result in legal liabilities, significant regulatory fines and irreversible damage to employer brand and customer trust. Many of these failures stem from relying on general-purpose AI tools. Because these models pull data from the open internet, they are highly susceptible to hallucinations and outdated information. To bridge this gap, forward-thinking organizations are moving away from broad tools and adopting specialized, domain-specific AI solutions. In HR and compliance, for example, utilizing AI trained on deeply vetted, compliant regional data, paired with a strict human-in-the-loop framework. This combination allows businesses to navigate complex local frameworks smoothly, ensuring they can scale globally with confidence rather than caution.