Block, the payments company behind Square and Cash App, announced it is laying off 40% of its workforce—4,000 workers—attributing the cuts directly to AI. Founder Jack Dorsey said AI tools have "changed what it means to build and run a company," and he was blunt about what comes next: "Within the next year, I believe the majority of companies will reach the same conclusion and make similar structural changes."
Dorsey framed this as a choice between honesty and delay—cutting gradually over months or years as AI adoption plays out, versus acting decisively now. He told employees the decision wasn't because Block is in trouble.
That framing matters, because a parallel debate is erupting over whether companies announcing AI-driven layoffs are being fully honest. Challenger, Gray & Christmas reported that employers cited AI in announcements of more than 50,000 layoffs in 2025, with Amazon, Pinterest, and Hewlett-Packard among them. But some experts suspect at least some of those attributions involve "AI-washing"—Forrester's term for attributing financially motivated cuts to AI implementation. The rationale, experts note, is "very investor-friendly" compared to saying the business is struggling.
The distinction matters for workforce planning. If cuts are genuinely AI-driven, that signals structural transformation is ahead for your industry. If they're AI-washed, the underlying business pressures are being obscured. Either way, Dorsey's warning isn't one HR leaders should dismiss. Read more via The Wall Street Journal, The New York Times, Gizmodo
This week's most dramatic AI story began with a $200 million Pentagon contract and ended with Anthropic designated a "supply chain risk to national security." Here's the sequence.
Anthropic held a contract making it the only AI company with a model deployed on classified military networks. The dispute ignited after Claude was allegedly used—without Anthropic's prior knowledge—during a raid that captured Venezuelan President Nicolás Maduro. Defense Secretary Pete Hegseth then gave Anthropic a hard deadline to grant the military unrestricted, "all lawful use" access to Claude. Anthropic refused, saying it could not allow Claude to be used for fully autonomous weapons or mass domestic surveillance, and that its models aren't reliable enough without human oversight. When the deadline passed, Trump ordered all federal agencies to immediately stop using Anthropic products. Anthropic called the action "retaliatory and punitive" and vowed to challenge the designation in court.
OpenAI subsequently announced a deal with the Pentagon to deploy its models on classified networks—though OpenAI CEO Sam Altman had previously said he agreed with Anthropic's position, prompting questions about how different the two arrangements actually are. Tech workers from OpenAI and Google signed an open letter supporting Anthropic's stance.
For enterprise AI buyers, the standoff raises a question worth asking your own legal and procurement teams: what are your AI vendors' red lines, and what happens when government pressure conflicts with those commitments? Experts say the standoff has set a potentially chilling precedent for how the government can pressure private AI companies. Read more via CBS News, CNBC, NPR, CNN
The Department of Justice has moved from signaling to investigating. A senior DOJ official revealed that tracking workforce demographic goals, connecting executive bonuses to diversity numbers, and requiring employees to create DEI-related goals that affect their pay or promotions are active enforcement priorities. Attorney Brenna Jenny of the DOJ's Commercial Litigation Branch specifically criticized what she called "color-coded charts" that treat race and sex as performance metrics rather than focusing on qualifications. When executive pay is tied to demographic targets, she said, "the practical pressure is obvious."
Notably, experts say the administration has achieved significant policy impact simply through the threat of enforcement—even without winning lawsuits in court.
But the picture is more complicated than a straightforward DEI retreat. Experts say many companies aren't genuinely abandoning diversity efforts—they're renaming them. "Equity" is being replaced with "belonging" and "inclusion," with the underlying programs largely intact. Employers are navigating a balancing act between avoiding federal targeting and avoiding boycotts and internal strife. Many high-profile announcements about dropping DEI are described by experts as "lightning rod, headline-generating activity" rather than substantive change.
Goldman Sachs added a concrete data point this week, removing race, gender identity, ethnicity, and sexual orientation from its board selection criteria following a demand from a conservative activist group with a small ownership stake. Goldman's governance committee will retain broader diversity factors like viewpoints, background, and work experience.
For HR leaders: the practical question isn't whether to publicly champion or abandon DEI—it's whether your specific practices (tracking demographic goals, tying exec comp to diversity numbers, building DEI into performance reviews) are now legally exposed. Read more via Bloomberg Law, Marketplace, The Wall Street Journal
Q4 2025 GDP growth came in at 1.4%—well below the 2.5% economists expected and a sharp drop from the 4.4% pace in the summer. For all of 2025, the economy grew 2.2%, the weakest since 2022. The Commerce Department estimated the government shutdown from October through November 12 cut roughly one percentage point from quarterly GDP growth.
Consumer confidence hit its lowest level in over a decade in January, with more Americans saying jobs were "hard to get." The Conference Board's Leading Economic Index logged its fifth consecutive monthly decline in December, with the LEI signaling "weaker economic activity at the start of this year." The Conference Board is projecting GDP expansion of 2.1% year-over-year in 2026.
On hiring structure: Goldman Sachs estimates the economy now only needs 50,000 new jobs per month to keep unemployment steady—down sharply from historical levels—because immigration enforcement has cut the flow of new workers by 80%, with net immigration projected to fall to just 200,000 in 2026. Job openings have already dropped below pre-pandemic levels to around 7 million. Read more via The Wall Street Journal, Fortune, The Conference Board
Long-term unemployment is becoming a "mental war." 25% of all unemployed Americans have been looking for work for six months or more—a share that has been rising for three consecutive years. Most have exhausted unemployment insurance. Job seekers tell CNBC the experience has eroded their confidence and that despite following every best practice, nothing is working. ZipRecruiter labor economist Nicole Bachaud notes that unemployment is "becoming more of a status quo versus a temporary position." Read more via CNBC, The New York Times
Walgreens is funding workers' path to becoming pharmacists. Facing a shortage severe enough that open pharmacist jobs now outnumber expected graduates, Walgreens launched PharmStart—now expanded nationwide—covering prerequisite courses and offering up to $40,000 yearly for pharmacy school in exchange for a four-year post-licensure commitment. Pharmacy school applications dropped by more than 64,000 between the 2011-2012 and 2021-2022 school years. Average pharmacy school debt exceeds $170,000. Read more via HR Brew
42% of tech workers now say their manager expects AI use daily. That's up from 32% just eight months prior, per new data. Google is factoring AI use into software engineer performance reviews for the first time; Meta is tracking AI-assisted lines of code; Salesforce now requires employees to use an AI agent to file for paid time off. Meanwhile, Elon Musk is bypassing resumes entirely for Tesla's AI chip team, asking applicants to submit three bullet points describing the hardest technical problems they've solved. Three-quarters of companies now use skills-based assessments during hiring, up from 56% the prior year. Read more via The Wall Street Journal, Fortune
Student loan delinquency has tripled since 2019. Nearly 25% of federal student loan borrowers failed to make required payments in 2025, up from less than 10% in 2019. Roughly 2 million borrowers experienced significant credit deterioration, with typical scores dropping 100 points. Racial disparities are pronounced: while 20% of white borrowers missed payments in Q3 2025, the rate among Black borrowers exceeded 48%. Read more via CNBC, The Century Foundation
Mark Cuban says there are two types of LLM users. Cuban divides AI adopters into learners—those who use it to understand everything—and non-learners, those who use it to avoid having to learn anything. He applies the same lens to companies: there will be those who are great at AI, and everybody else. (Business Insider)
Amazon is blaming AI for recent AWS outages. Two recent Amazon Web Services outages occurred as a result of actions by Amazon's own AI tools, per recent reports. Amazon's AI agent "Kiro" was reportedly responsible for a December incident affecting an AWS service in mainland China, having chosen to "delete and recreate the environment" it was working on. This is the second outage linked to an AI tool in recent months. (The Verge)
PwC has a new AI agent built for enterprise spreadsheets. While traditional AI tools "just kind of shrug and give up" when confronted with large spreadsheets, PwC's new frontier AI agent is designed to navigate them the way experienced practitioners do—scanning, searching, jumping across tabs, integrating charts and receipts, and reasoning across the data. (Business Insider)
OpenAI considered—then declined—to report a mass shooter. In the months before Jesse Van Rootselaar became a suspect in a British Columbia mass shooting, OpenAI employees considered alerting law enforcement about posts he made on ChatGPT describing gun violence scenarios. Those posts were flagged by an automated review system. OpenAI ultimately decided not to contact authorities. (The Wall Street Journal)
Canada: Manufacturing activity fell in January, with factory shipments down 3.3% from December—the third decline in the last four months. Steepest drops were in transportation equipment and machinery. (The Wall Street Journal)
Greece: Lawmakers are advancing legislation that simultaneously recruits tens of thousands of South Asian workers while penalizing undocumented migrants—a balancing act between projecting toughness on immigration and addressing a serious labor shortfall in tourism, construction, and agriculture. (Politico)
Ireland: Early evidence suggests AI is weakening employment opportunities in parts of Ireland's technology-focused economy, particularly for young graduates. Employment in high-AI-risk sectors like tech and financial services grew 4% from 2023 to 2025, compared to 6.25% growth in low-risk categories over the same period. (Reuters)
Italy: Unemployment held stable in December at 5.6%, below economists' forecast of 5.8%, though the labor market shed a net 20,000 jobs during the month. Youth unemployment rose from 19.1% to 20.5%. (Reuters)
Switzerland: GDP expanded 0.2% in Q4 2025 after contracting 0.5% in Q3, with the economy growing 1.4% over all of 2025—faster than 2024's 1.2% but still below Switzerland's average growth rate. (The Wall Street Journal)
United Kingdom: Job vacancies fell to their lowest level in five years in January, with 694,940 advertised positions—a 16% year-over-year decline and the first time vacancies have dropped below 700,000 since January 2021. (SIA)
Get the full interactive edition → View in digital reader
Stay ahead of the market → Subscribe to the Need to Know Briefing to get workforce intelligence delivered to your inbox every week.
The Need to Know Briefing is published weekly by Kelly, curating the most important workforce and hiring insights for HR leaders and hiring managers.