From AI “wild west” to enterprise guardrails, the organization behind the Grammys is using Anthropic’s Claude to fast‑track workforce strategy and reset talent expectations
Shonda Grant, Chief People and Culture Officer at the Recording Academy, is steering a dual-track transformation that spans both HR operations and enterprise-wide change. At the organization behind the Grammys, AI adoption has become a central lever for scaling efficiency and modernizing workforce strategy. The shift reflects a broader reality facing large institutions: technology is advancing faster than organizational frameworks can keep pace.
Grant described an environment where experimentation and urgency coexist, driven in large part by executive alignment. “Change is really coming from the top,” she said, pointing to strong support from both the CEO and president in advancing AI initiatives. That backing has enabled the Academy to move quickly from informal usage toward structured deployment.
Building guardrails in an AI ‘wild west’
The organization is currently piloting an in-house version of Claude, developed by Anthropic, as part of a broader effort to formalize AI usage. The pilot, led by IT in collaboration with strategy and executive leadership, includes approximately 20 users drawn largely from senior leadership. This controlled rollout is designed to test practical applications while gathering feedback on risks, proper governance, and productivity gains.
Grant framed the current moment as analogous to the early days of social media adoption in corporate environments. “It was sort of the wild west,” she said, noting that employees are already using tools like ChatGPT, Gemini, and Claude independently. Rather than restricting usage, the Academy is focused on establishing guardrails that align with organizational goals and data security requirements.
That shift has required rapid policy development, particularly around data handling. Initially, employees were instructed not to input proprietary information into public AI tools. However, the introduction of an internal platform changes that calculus, allowing for more secure use of sensitive data within defined boundaries.
“It’s hard to keep up with how rapidly the technology is evolving, but we want to be an early adopter in the AI transformation that’s currently taking place,” Grant said. The pace of innovation is forcing organizations to adopt iterative oversight models rather than waiting for fully mature frameworks.
From transactional HR to strategic enablement
Within the People & Culture team, the implications are particularly pronounced. Grant emphasized that AI is not being positioned as a replacement for employees but as an enabler of higher-value work. “The goal is for us to be more efficient, to streamline our manual processes,” she said.
This aligns with a longer-term and broader overall shift of the HR function in many organizations, from administrative or transactional functions to more strategic roles such as leadership development, organizational design, and coaching. By automating routine tasks, AI creates capacity for HR professionals to focus on initiatives that directly influence business performance and culture.
At the same time, workforce concerns remain a persistent challenge. “People are worried that AI is going to replace them,” Grant said, acknowledging that apprehension exists even at times within her own team. Addressing those fears requires clear communication about intent, as well as visible investment in upskilling and career development.
The emergence of agentic AI—systems capable of executing tasks autonomously—adds another layer of complexity. Grant described it as both “exciting and terrifying at the same time,” reflecting the dual nature of its potential impact. While agentic systems could significantly augment productivity, they also raise questions about workforce structure and capability requirements.
Organizations are already under pressure to deliver more with constrained resources; a trend Grant sees as universal across every industry. “We’re never going to be asked to do less,” she said, noting that initiatives are more often added than removed. In that context, AI becomes a critical tool for sustaining output without proportional increases in headcount.
Hiring, upskilling, and the race to readiness
Looking ahead, Grant identified workforce readiness as a defining challenge of the next phase. This includes both hiring talent with existing AI expertise and rapidly upskilling current employees. The bar for new hires is rising quickly, reflecting the accelerating integration of AI into core business functions.
“If we’re hiring people that say, ‘I really want to learn about AI,’ then they’re already too far behind,” she said, relaying guidance and expectations from the organization’s CEO. The bar is shifting toward candidates who arrive with practical AI fluency and can immediately apply those capabilities.
For existing staff, the focus is on empowerment through training and access to tools. People & Culture’s role is to ensure that employees are not only equipped to use AI but also understand how it reshapes their roles and responsibilities. This requires continuous learning models rather than one-time training interventions.
The speed of change, however, complicates even well-intentioned efforts. “It’s nearly impossible to have all that in place before it hits,” Grant said, highlighting the challenge of building policies and training programs in parallel with rapidly evolving technology.
Change management remains a critical barrier, particularly among leaders who may be skeptical of AI-driven workflows. Grant stressed that mindset shifts are as important as technical adoption. “There are some leaders who are hesitant about staff using AI to assist them in doing their work,” she said.
Top-down alignment is therefore essential. Organizations without strong executive sponsorship are likely to lag in adoption, as cultural resistance slows implementation. At the Recording Academy, leadership support has enabled a more proactive approach, positioning AI as a strategic priority rather than a peripheral experiment.
