There’s been plenty of discussion lately over AI’s potential impact on jobs across industries. Which positions will survive? Which ones could be eliminated over time? And what new types of AI experts should companies be seeking out in this time of dramatic change?

Less discussed, until now, is how talent and people leaders at organizations today should be managing and thinking strategically about all this inside their rapidly transforming organizations.

Over two recent, in-person workshops, we gathered people and talent leaders from across the Battery portfolio and our broader network to compare notes on this topic and on navigating the AI era generally. After our discussions, we realized it’s really one of the more disorienting moments in modern workforce history. AI is changing the way we’re thinking about hiring, compensation, career paths, performance management and more. It’s also made clear that HR can no longer operate as a reactive function. The companies navigating this era most effectively aren’t treating their people leaders as policy enforcers or process managers; they’re treating them as architects of how the business runs, and the smartest people leaders aren’t waiting to be invited into that role.

Hiring has (surprise) become more deliberate and competitive

With many companies replacing roles or entire departments with smart AI agents, it’s no surprise that open seats at organizations are no longer automatically backfilled.

As companies evaluate vacancies and how to fill them today, many leaders are engaged in proactive market mapping (think researching and charting the external talent landscape within a specific industry, skill set, or geography) and pulling in internal leaders and individual contributors for their input before starting to fill a role.

A concept that kept surfacing in our discussions was “KYC”, or “Know Your Candidate.” As competition for top AI talent intensifies, recruiting teams are adapting this idea from the “know-your-customer” mantra in financial services to talent: Go deeper on what actually motivates a candidate so you can close them when it counts.

Some companies, for example, are getting to KYC by asking pointed questions around candidates’ AI fluency, like, “What tools are you using and how are you using those tools to accelerate business outcomes?” They’re also posing questions about AI usage and probing what KPIs candidates are implementing to understand their team’s usage of increasingly pricey tokens/compute. They might ask: “What are your thoughts on ‘tokenomics’ and how are you assessing that as your organization grows?”  All candidates are, by default, expected to be at least somewhat fluent in AI.

AI is also changing the recruiting process itself. Virtual career fairs, conversational hiring events and AI-driven, resume-matching tools are becoming standard parts of the pipeline, not just to save time, but to find candidates who might be good fits. Many companies are also raising referral bonuses to compete for a shrinking pool of high-impact candidates, even paying 70-80th percentile of bands for key hires.

Can you grow without growing headcount?

This topic sparked the most debate. But the question itself is a critical signal: Companies that are asking “do we hire or do we automate?” as a binary choice are usually asking too late—after a role has opened, or after a team is already stretched. It’s preferable to get ahead of that question by doing the harder work of mapping how work flows through the organization before deciding which resource should do it.

That means taking workflow design seriously as a strategic exercise. The companies getting this right are sitting down with functional leaders and walking through their core processes end-to-end. They’re asking where time is really going, where handoffs break down, where decisions get made and by whom, and where productivity drag lives.

The takeaway that resonated most in our discussion: Before you open a req, map the workflow. Understand what the role truly entails at the task level, which parts of that work could be augmented or automated, and what you’re hiring for once you strip those parts away. That exercise often changes the job description entirely and sometimes eliminates the need for a hire, redirecting budget toward tooling.

This is also why job expectations themselves are shifting visibly. Employees are increasingly expected to be part functional expert, part systems thinker and part AI operator. People at all levels need to do their core job and, at the same time, help redesign how work gets done with AI. That’s a meaningfully different profile than what companies were hiring for a few years ago, and it’s why traditional job-description language feels stale. Roles like “AI enablement lead” or “GTM engineer” didn’t exist in their current form a few years ago but are critical in many organizations now. More broadly, companies are also grappling with how to turn pockets of AI progress in various functions into company-wide AI strategies.

Pay is getting more targeted—and candidates are opting for “rock-climbing walls” over ladders

The “peanut butter” approach to compensation—spreading raises evenly across the org—is losing ground, according to our portfolio’s people leaders. Companies are concentrating resources on the roles and people that create the most leverage and building compensation structures that reflect that. The message from the room was consistent: Reward impact, not tenure.

Retention is a more nuanced story. Money matters, but for a growing number of employees it’s less important than opportunity. A great metaphor emerged in the workshop from our friends at Sequoia: The companies winning the talent war are offering employees a rock-climbing wall, not a corporate ladder.

A corporate ladder tells you exactly where you’re going and exactly how long it will take to get there. Progress is linear, predictable and slow. A rock-climbing wall is different: There are dozens of routes to the top, some unexpected, and the most interesting climbers are the ones who find a path nobody has mapped before. The wall rewards people who are curious and adaptable, not just patient.

A newer generation of employees are less interested in a guaranteed trajectory and more interested in building a portfolio of experiences that compound over time. Startups have always operated this way. What’s new is that they’re explicitly positioning the breadth and speed of the experience as a feature, not a consolation prize for lower pay. For companies that can’t compete on salary alone, that framing is increasingly a legitimate edge.

Equity structures are evolving as well. As more late-stage companies stay private longer, the traditional stock-option model looks more outdated. Restricted stock units (RSUs) are gaining traction as an alternative, particularly at companies where rising strike prices have made options feel less meaningful to newer hires. Unlike options, RSUs don’t require employees to pay anything to receive them; once they vest, they’re simply granted as shares. That predictability is increasingly appealing both to employees who want to know what they’re getting, and to companies looking for cleaner retention tools that don’t depend on a specific liquidity outcome. Our deep dive on RSUs is a useful primer on when and why making that switch makes sense—including the real trade-offs leaders need to consider before doing it.

Refresh equity is getting more selective in this environment, with one notable exception: AI companies are still extending refresh grants broadly, because the competition for that talent hasn’t let up. For everyone else, the retention playbook is leaning harder on what money can’t easily replicate—the sense that you’re growing, that your work is visible, and that the next two years will teach you something the next company couldn’t.

The people leader’s mandate is changing—and the window to act is now

The decisions being made now about AI, headcount, org design and compensation aren’t just operational decisions but strategic ones—and they’re going to shape how companies compete for the next decade. People leaders’ experience in areas like compliance, process and risk management is more important than ever in today’s AI world, and these skills empower forward-thinking people and talent leaders to play larger, more strategic roles within their organizations. That is where they’ll drive the most strategic impact and move their companies forward in the AI age. What was once a seat at the table is becoming a hand on the wheel.