Newsletter Archive

Browse past editions of Work Edge. Each teaser shows The Hook and The One Big Thing. Full editions and archive access are included with a paid subscription.Read a full edition free Get The Radar free

Newsletter Archive

The most important AI story this week isn’t about a product launch or a funding round. It’s about what happens when AI quietly absorbs the connective tissue of how organizations actually run.

The market just handed every CFO a loaded question for your next board meeting, and "we're still assessing" is not going to cut it.

When McKinsey is your AI vendor's sales rep and a coding agent fits in a developer's backpack, the question isn't whether autonomous AI is coming to your workforce — it's whether you'll define the terms before someone else does.

The governance question your org hasn't answered yet became a 13-hour outage last week, and the agents causing that kind of damage are already inside your systems.

The most important governance story of the week didn't come from a regulator or a court. It came from Anthropic's own data, and it shows that the oversight model most enterprises built their AI policies around is eroding in real time.

Spotify's best engineers haven't written a line of code in 2026 — and 72% of your workforce still thinks AI won't touch their job.

Microsoft's AI CEO just put a timestamp on white-collar automation—12-18 months—while Spotify's best engineers haven't written code manually since December and a Chinese AI lab says 30% of its daily work across HR, finance, and sales is already automated.

Harvard research tracking 200 employees found AI tools increased workloads rather than lightened them—raising urgent questions about whether your AI investments are boosting productivity or accelerating burnout.

SHRM research reveals 244,000 HR jobs—24% of the function—are now majority-automatable by generative AI, which means CHROs face the uncomfortable task of auditing their own teams before advising the business.

When AI can autonomously complete six-hour engineering tasks, the workforce planning question shifts from "which tasks can we automate" to "which projects still need humans."

Deloitte is retitling 182,000 employees to reflect AI-augmented work by June—the clearest signal yet that traditional job architecture is becoming obsolete.

SHRM research finds 393,000 HR jobs face 50%+ automation risk—meaning the function leading workforce transformation is now the workforce being transformed.

Companies are laying off workers based on AI's *potential* rather than its proven capabilities—while only 3,000-5,000 people globally can actually implement AI that delivers value.

SHRM just mapped AI displacement risk across 10 HR occupations—and for the first time, HR has to confront that the function advising everyone else on workforce transformation is itself on the transformation list.

Anthropic's CEO just put a number on AI displacement—50% of entry-level office jobs at risk within five years—and the UK is already showing 8% net job losses, giving CHROs the first hard benchmarks to stress-test workforce plans against.

Architects are redesigning offices around AI agent noise—which means your workforce planning is already behind where facilities teams assume you'll be.

Fortune 500 CEOs are standing on the Davos stage naming customer service and programming as jobs that "will go away"—while AI agents just built a working web browser in one week.

Anthropic just released the most concrete AI workforce data we've seen—AI now handles 25% of tasks in nearly half of all jobs, with 12x speedups on college-level work—and if your organization isn't seeing those gains, you have an adoption problem, not an AI problem.

AI agents just built a 3-million-line web browser from scratch in under a week—your engineering headcount model was designed for a world that no longer exists.

McKinsey now counts 25,000 AI agents as workforce members alongside 60,000 humans—and your board will soon ask what your ratio is.

Anthropic just shipped AI that executes entire workflows autonomously—while new benchmarks show even the best models hallucinate on half of real work tasks, meaning your "AI strategy" now requires both acceleration and guardrails simultaneously.

Southwest's CEO just admitted AI tools work fine—it's getting 72,000 employees to actually use them that's the problem, and that confession should reframe every AI initiative on your 2026 roadmap.

Utah just gave an AI system the legal authority to prescribe medications without a doctor's approval—and a dozen more states are watching, signaling the end of the assumption that licensed human judgment is irreplaceable.

Companies are laying off workers based on AI's potential rather than its proven capabilities—while only 3,000-5,000 people globally can actually implement AI that delivers value.