Layoffs Are Holding Steady, but the Jobs You’re Posting May Already Be Out of Date
Sign Up For Daily Job Market News
First-time jobless claims barely moved last week. The U.S. Department of Labor reports that 203,000 people filed new unemployment claims in the week ending November 14, up 1,000 from the week before and below the 222,000 a year ago. The 4-week average, which smooths out the weekly bumps, rose 2,000 to 197,750, compared with 224,750 a year ago. Initial claims are 19,000 below a year ago, and the 4-week average is 27,000 below.
The bigger move came in continuing claims, the count of people still collecting benefits. They rose 70,000 to 1,763,000 for the week ending November 7, the largest one-week increase in the release’s 12-month table and the highest level since the week ending August 29. Continuing claims are still about 186,000 below the 1,949,000 from a year ago, and the insured unemployment rate held at 1.1%.
Most of the raw increase traces to one state. Before seasonal adjustment, continuing claims rose 42,996 nationwide, and Michigan alone rose 39,343, from 30,078 to 69,421. Michigan’s new claims also jumped 6,095 to 8,473 in the latest week, and the release gives no reason. Other states did explain their larger increases in new claims for the week ending November 7: education layoffs in New Jersey, manufacturing in Missouri, manufacturing plus transportation and warehousing in Texas, and construction in Minnesota.
Florida stayed quiet. New claims there fell 661 to 4,701 in the advance count, and the state’s insured unemployment rate was 0.3% for the week ending October 31, against 2.0% in New Jersey, the highest.
The headline number is boring, which is good news. New claims sit below a year ago, so employers are not cutting people in any real volume. The 70,000 jump in continuing claims would worry me if it showed up across the map. Almost all of the raw increase came from Michigan, and Michigan moved the other way the week before, down 14,771. A drop like that followed by a spike looks like a one-state event, and I want to see next week’s number before I read anything national into it.
For hiring managers, the main read has not changed. Continuing claims are still well below last year, so a hiring plan that counts on laid-off workers filling your openings has a smaller pool to draw from. If you need to fill roles, explore our staffing services and we can start with the people you cannot reach on your own.
Staffing your team doesn’t have to be hard.
Reach out and see how we can help.
Banner Ads and Backlinks Are Fading as Upwork Clients Hire for AI Video and AI SEO
Upwork’s Research Institute released its monthly hiring report today. It covers hired contracts from U.S. clients on Upwork’s freelance marketplace, comparing October with September and with October 2025. The report ranks skills and counts growth streaks and publishes no contract counts or percentages, so none appear here.
Marketing skills show the clearest split. “AI Ad Creation” and “AI Animation Tools” each grew for 3 straight months, and “AI Video Ads” grew for 2. “Banner Ad Design” fell from both September and a year ago. Search marketing looks the same. “AI SEO” grew for a second straight month, while “SEO Backlinking” fell month over month and year over year. Upwork ties the shift to people using AI chatbots and Google’s AI Overviews to find answers.
Client searches point the same way. “AI video” was the most-searched AI term for a second consecutive month, and “AI developer” ranked second. “AI UGC video,” meaning AI-made video styled like creator content, was the fastest-growing AI search compared with last year. “AI agent developer” and “AI SaaS developer” ranked second and third for growth.
Software testing was the other standout. “Software Testing,” “Functional Testing,” and “Usability Testing” ranked second, third, and fourth among skills growing both month over month and year over year, with “Unreal Engine” first and “Claude” fifth. Upwork says its data does not link the testing contracts to AI projects.
Some skills cooled. “Data Annotation” and “Data Labeling” stayed above last year’s levels but slipped from September. “Android” and “iOS” picked up from September and remain below a year ago.
Jennifer Brett, managing director of the Upwork Research Institute, says, “A marketing role may keep its title while the work changes.” Her line is the line for hiring managers to remember. If your marketing job description still asks for banner design and link building, you are describing work the market is moving away from. I covered the Upwork and OpenAI research on how AI tasks pair with traditional work, and this report adds a month of hiring evidence to it.
I see the SEO piece firsthand. Getting cited in AI answers is now part of the job for anyone responsible for a company website, and the skills behind it look different from classic link building. I tracked which AI skills employers were searching for on Upwork in October, and the pattern continues.
The testing result interests me as much as the marketing one. Testing skills hold 3 of the top 5 spots among skills growing both month over month and year over year, which tells me product quality is an active area of hiring. I would call a good tester worth recruiting right now, and I would not read more into the AI connection than Upwork does.
Employers Blame Job Security Fears for Hidden AI Use, but Only 24% of Employees Agree
A survey from work-management software company Slingshot found a gap between how much employees use AI at work and how much they tell their bosses.
Slingshot surveyed 500 full-time U.S. employees and managers across 4 age groups. 45% of employees keep their AI use at work private to some extent, yet 60% of employers believe their employees are fully open about it. The page does not say how many of the 500 were employers, so the employer figures may rest on a small group.
The two sides disagree on why. 47% of employers think job security fears drive the secrecy, but only 24% of employees say so. The top employee reason, at 45%, is that they do not think they need to disclose it, and only 16% of employers guessed that. Another 34% of employees worry others will think they are cutting corners, and 27% fear being judged.
Gen Z feels the most pressure. 47% of Gen Z workers cite fear of being judged, and 44% worry about looking like they are cutting corners. Millennials (44%), Gen X (57%), and Boomers (47%) mostly say they do not feel obligated to share.
Most of the use is voluntary. Only 28% of businesses require employees to use AI tools, while 87% of employees use it on their own. Employees mostly use AI to check or improve their work (54%) and to draft emails and reports (52%). Managers use it to analyze data (56%), do research (52%), and manage team priorities (47%).
The gap on motive is what jumps out to me. Employers assume fear, while most employees simply do not think anyone needs to know. My read is that a missing company policy leaves that decision to each employee. If your company cares whether people use AI, put it in writing. Tell people which tools are approved and what they need to disclose. Without a written policy, your newest employees are the ones most worried about getting judged.
For job seekers, use AI to do better work, then say how in your interview. A project where you used AI, checked its work, and fixed what it got wrong is easy to explain. See what’s open right now.
