• The Big Shift: AI @ Work
  • Posts
  • States push back on AI rules freeze, Big Tech flexes at work, and hybrid gets redefined by bots

States push back on AI rules freeze, Big Tech flexes at work, and hybrid gets redefined by bots

Author

Henry Nothhaft, Jr.
May 24, 2025

Your go-to rundown on AI’s impact on the future of work—delivered every Friday. Each edition highlights three to five must-read stories on everything from job disruption and upskilling to cultural shifts and emerging AI tools—all in a crisp, Axios-style format.

Was this newsletter forwarded to you? Sign up to subscribe.

In the wake of the week…

We may be a day late with this week’s digest, but we’re certainly not a dollar short.

This week begins with the escalating battle over who sets the rules for AI. While enterprise leaders push for responsible, human-centered adoption, lawmakers are debating whether states should retain the authority to regulate at all. Google and Microsoft, meanwhile, are flexing their business models and platform power, doubling down on efforts to prove who can best deliver AI where work already happens. And finally, we return to the evolving definition of hybrid work, not as a question of geography, but as a model defined by collaboration between people and intelligent systems.

And in this week’s Extra Credit: UBS turns analysts into avatars to scale client insights, AI agents hit a final technical hurdle, and executives use copilots to conquer the post-vacation inbox—and take back their time.

And a quick plug: Tomorrow I’ll publish a Sunday Prompt that takes a deeper dive into themes we’ve been tracking in this digest for months. AI at Work: Displacement, Power, and What Comes Next looks harder at where things are heading and what’s at stake. I hope you’ll give it a read.

Now let’s dive in. 👇

AI Regulation Enters a Critical Crossroads: Business, Ethics, and Federal Power Collide

A two-front battle over the future of responsible AI is unfolding inside the enterprise and across Capitol Hill. While leaders like strategist Alison McCauley call for thoughtful adoption rooted in human values and strategic learning, federal lawmakers are advancing legislation that would block states from enacting their own AI safeguards for a decade.

McCauley, an early AI advisor, frames the current moment as a “critical learning window.” She argues that generative AI changes how business gets done at a structural level and demands new modes of human–machine collaboration. Responsible deployment, she says, begins with upskilling, ethical framing, and deliberate experimentation. She urges leaders to center human insight and agency, not replace it.

But just as enterprise leaders begin investing in that future, political developments threaten to undermine the patchwork of state-level protections that have served as the frontline of AI regulation in the U.S. A proposed 10-year federal moratorium would override more than 130 active or pending state laws, including California’s AI disclosure and privacy statutes. Critics warn it would create a legal vacuum just as AI systems are being embedded into hiring, healthcare, education, and finance.

By the Numbers

  • 600+ state-level AI bills introduced in 2025 (Transparency Coalition)

  • 42 AI-related laws passed in California from 2015–2024 (Stanford AI Index)

  • $25 million projected cost of Texas’ AI regulatory framework (Texas Tribune)

  • Only 4 federal AI bills passed in 2024, out of 220+ introduced (AI Index)

Why It Matters

As AI scales across industries, the gap between innovation and governance is widening. Executives face increasing pressure to adopt generative tools while navigating ethical, regulatory, and reputational risks. A federal freeze on state regulation would remove hard-won protections without offering a coherent national alternative, leaving organizations and consumers exposed.

Leadership Insight

The regulatory environment is uncertain, but responsibility cannot wait. Executives should not treat AI ethics as a compliance issue. It is a trust issue, a brand issue, and increasingly, a business risk issue. Proactively adopting responsible AI frameworks and surfacing internal governance structures can create stability amid legal ambiguity and demonstrate leadership to customers, investors, and regulators alike.

The Bottom Line: AI is reshaping business faster than regulators can keep pace. Whether or not Congress preempts state action, companies must take the lead on responsible adoption. The costs of delay—legal, ethical, and reputational—are rising by the week.

Google and Microsoft Invest Heavily in AI Catch-Up, Leveraging Enterprise Clout

I recently argued that the AI race will ultimately favor incumbents like Google and Microsoft, leveraging their existing platforms, enterprise relationships, and deep integration capabilities. This week's developer conferences from these tech giants began to validate that prediction, demonstrating their significant investments to catch up with early movers OpenAI and Anthropic by flexing their considerable enterprise influence.

At Build, Microsoft emphasized enterprise-focused AI agents, unveiling tools designed to automate workflows securely and efficiently. The introduction of Microsoft's AI Gateway and identity management systems reflect strategic moves aimed at meeting enterprise needs for security, governance, and integration. An accidental disclosure revealed Walmart's preference for Microsoft's AI security solutions over Google’s, highlighting Microsoft's advantage in enterprise trust and compliance.

Google’s I/O event showcased deep integrations of generative AI into its Workspace suite, enhancing productivity by automating routine tasks across Gmail, Docs, and Vids. Additionally, Google's announcement of high-end subscription tiers reveals a strategic shift toward usage-based business models to sustain future growth, differentiating itself from traditional ad-driven revenue streams.

By the Numbers

  • Google’s Gemini models now process over 480 trillion tokens monthly, a 50-fold increase from last year (Tokens are the raw units of text AI models process—more tokens means more usage, more interaction, and greater model demand.)

  • Microsoft's AI agents saw daily active users more than double compared to the previous year

  • Google's AI queries can cost up to $3,500 per complex request (Business Insider)

  • Enterprise AI adoption is expected to grow 327% by 2027 (Salesforce)

Why It Matters

Google and Microsoft’s aggressive AI investments highlight the importance of platform integration, enterprise relationships, and business considerations beyond pure technological innovation. Microsoft's strong focus on security and governance addresses key enterprise pain points, while Google's productivity-driven enhancements align closely with daily enterprise workflows.

Leadership Insight

Enterprise leaders should prioritize AI solutions that integrate easily into existing workflows and infrastructure. Google's and Microsoft's recent moves demonstrate that effective AI deployment hinges on robust integration, data governance, security, and clear alignment with enterprise workflows and strategic objectives.

The Bottom Line: As predicted, Google and Microsoft are rapidly positioning themselves as major players in enterprise AI by leveraging strategic integration and substantial enterprise clout rather than focusing solely on standalone product innovation.

AI Layoffs Backfire as Hybrid Work Takes on a New Meaning

The world of AI moves fast. In just the few short months I have been publishing this newsletter, we have watched a near full arc play out. Companies, especially in Silicon Valley, which often serves as the sandbox for workplace experiments, charged headfirst into AI-first, worker-light strategies. The technology was not fully matured, and the culture around it was unprepared. Many are now revising course. A new hybrid model of work is taking hold, defined not by location, but by collaboration between human and machine. You can probably guess what comes next.

Orgvue’s latest research reveals a growing regret among business leaders who used AI as a justification for workforce cuts. Over half of those who eliminated roles due to AI now say it was the wrong move. At the same time, thought leaders and analysts are signaling a new model of “hybrid work” not defined by location, but by the interplay between humans and intelligent systems.

The data shows that companies expected rapid ROI from AI deployments. Instead, they found talent gaps, stalled productivity, and cultural backlash. While 72% of HR leaders still believe AI is the dominant driver of workforce transformation, most also admit they underestimated the complexity of implementation and overestimated AI’s ability to fully replace people.

By the Numbers

  • 39% of global leaders have made redundancies due to AI (Orgvue)

  • 55% of those leaders say they now regret the decision

  • 72% of HR leaders say AI is the top driver of workforce transformation

  • 80% plan to reinvest in AI in 2025, despite uncertain ROI

Why It Matters

AI was pitched as a massive efficiency unlock, but companies are learning the hard way that execution trumps theory. Firing staff before redesigning workflows, retraining teams, or mapping integration points leads to regret. The emerging consensus is that hybrid roles, part human, part AI, offer more sustainable business value than wholesale automation.

Leadership Insight

Do not treat AI as a shortcut to headcount reduction. Instead, build an AI adoption strategy around collaboration, augmentation, and trust. Equip teams to use AI effectively. Redesign roles to take advantage of what each party does best. And measure success in terms of performance lift, not payroll cuts. And do not underestimate the importance of transparent communication and cultural readiness. Successful transformation requires both.

The Bottom Line: AI is not yet a full substitute for human talent in knowledge work. It is a partner to it. Companies that treat AI as a collaborator, not a driver of staff reduction, will realize greater gains and avoid costly missteps.

Extra Credit

For the overachievers: These are the stories that didn’t crack the top group this week but are too important to ignore—quick hits on what’s happening and why it matters.

AI Agents Face Fundamental Hurdles

Key Takeaway: AI agents are advancing fast, but cannot fulfill their promise, like booking flights or managing schedules, until they are granted secure, authorized access to third-party apps and systems.

Why It Matters: Authorization may determine how quickly AI agents move from novelty to utility. Standards like the Model Context Protocol and startups like Arcade.dev are laying the groundwork. Once solved, the door opens to a wave of fully autonomous digital taskworkers embedded in everyday life.

Source
Wall Street Journal: AI Agents Face One Last, Big Obstacle 

AI Becomes the Executive Assistant for the Post-Vacation Catch-Up

Key Takeaway: Executives are increasingly turning to AI tools to manage the stress of reentry after time off, using copilots and custom agents to summarize Slack threads, triage email, and surface key decisions without relying on staff for manual updates.

Why It Matters: This use case highlights AI’s evolving role in knowledge work, not as a replacement for people, but as a productivity scaffold that reduces friction and restores time. As more leaders adopt these tools, the concept of time off may finally evolve to include real detachment, without operational debt.

Investment Bank UBS Taps AI to Create Analyst Avatars

Key Takeaway: Investment banking powerhouse UBS is turning research notes into AI-generated videos using avatar versions of its analysts, enabling more scalable, client-facing content without overloading staff.

Why It Matters: As client expectations evolve, especially in high-touch sectors like finance, AI is not just streamlining workflows—it is reshaping how professional expertise is packaged and delivered. UBS’s move illustrates how firms can use AI to personalize and scale without compromising human oversight.

This edition of The Big Shift: AI @ Work may have been edited with the assistance of ChatGPT, Claude, Copilot, Gemini, Perplexity, or none of the above.

Want to chat about AI, work, and where it’s all headed? Let’s connect. Find me on LinkedIn and drop me a message.