Tech layoffs keep coming in waves, and the startup data is thinner than the headlines
US tech layoffs and restructuring rolled into early November. Comcast filed state-level layoff warnings on 2 November, and coverage of Big Tech's repeated job cuts continued the same day. Hard startup-specific data is thin.

The newest items on tech job cuts carry dates, not numbers. Comcast's state-level layoff warnings began to arrive on 2 November. A CNBC piece the same day described Big Tech's job cuts as arriving in waves, with workers laid off and then laid off again. Neither item appears as full text, only as a headline. That matters for anyone reading the surge of 2026 layoff coverage as a startup story.
The strongest primary material on technology and work is not a layoff tally at all. It is a research agenda. A Knight First Amendment Institute paper, "AI as Normal Technology," argues that AI's economic effects will play out over decades, not quarters. It says the useful distinction is between AI methods, AI applications and AI adoption, which move on different timescales. The authors say they have not tried to quantify probabilities. They describe what they see as the median outcome instead. That framing sits awkwardly beside the way layoff news is usually written. A single quarter's cuts get presented as proof of an AI-driven restructuring, when the paper argues that adoption is slow, uncertain and shaped by institutions.
The Knight authors also reject the idea of AI as a highly autonomous entity. They call that framing neither accurate nor useful for understanding societal impact. Read the paper, and the wave metaphor in the headlines starts to look like a description of press cycles.
Citations without a denominator
One real number does come from the laboratory of AI work. An estimate says AI data labeling and edge-case work is poorly paid, boring, and may not be temporary. A Computational Complexity blog post, published on 4 October, makes the case. People in AI once assumed such jobs would soon be automated, the post says. Now they suspect they will not be, because the more we expect from AI the more edge cases appear.
The post is explicit that its source is an article it links to, not original reporting. It was written about nine months before publication. It is context, not a current headcount. No startup-specific layoff figure dated inside the last 72 hours appears here. The context headlines mention Pipedrive's "major" restructuring on 28 October and a voluntary exit programme at Moniepoint covering engineering and product on 27 October. Those are supplied as context only and are not citable facts under this desk's rules. Any piece that puts a startup number in the lead this week is drawing on something else.
"Whether or not consumers are aware, big tech companies are all over the vehicles that we drive," said Nicole Zagson, a doctoral candidate in cybersecurity at Northeastern and a study co-author.
The nearest thing to a hard, datable finding about technology and its users comes from Consumer Reports. It published on 4 October a study run with privacy researchers at Northeastern University. The team tested 21 vehicles and their companion apps inside a Faraday tent at CR's Connecticut auto test centre. The cars were model years 2022 through 2025, many of them EVs. Almost a quarter of the vehicle apps were sending out personally identifiable information, according to the report. That included owners' names, vehicle identification numbers and precise geographic locations. The top recipients included Amazon, Google, Meta, Microsoft, Pinterest, Snap and Yahoo. It is not a restructuring story, but it is a dated data point about where value is being extracted in tech.
The same report quotes another co-author, Sarah Elizabeth Gillespie: "It does not appear that a customer can buy a new car that does not track you." Several automakers told CR that third-party companies can embed pixels and cookies on webpages opened from their connected apps. The study itself is due to be published this week, per CR.
Product bets, not headcount
The startup news points in a different direction from cuts. TechCrunch reported on 3 October that Neko Health has opened in New York. The body-scan startup was co-founded by Spotify's Daniel Ek and Hjalmar Nilsonne. The company has raised nearly $1 billion total and says its global waitlist tops 300,000 people. A one-hour scan costs $500, is not covered by insurance, and should qualify as an expense for health savings accounts, according to the report. Farooq Abbasi of Preface Ventures, an investor, told TechCrunch that preventative care should be cheap and accessible. His own scan measured 55 biomarkers, up from 15 in his generation. That is expansion, not retrenchment.
The same is true of the Chick-fil-A story. CEO Andrew Cathy told CNBC, as reported by Fox Business on 4 October, that the chain will not use AI to take drive-thru orders. "From our experience, we really want that hospitality to be human to human," Cathy said. "We're not going to substitute that interaction with technology." McDonald's said at its investor day last month that it plans to test voice AI called Archy in English and Spanish. Wendy's has been deploying its FreshAi system built with Google Cloud. Chick-fil-A's choice preserves human jobs at the order point. It is a strategy decision, not a restructuring.
The clearest restructuring here is a political one. It concerns government technologists rather than startups. Writing on the Can We Still Govern? Substack on 4 October, Merici Vinton says the Federation of American Scientists and Georgetown's Beeck Center ran a series of retrospectives. More than 100 digital government practitioners took part.
She describes the field as having undergone a massive rupture. The US Digital Service became DOGE and 18F shuttered, while other agency-level digital teams were closed or reshaped. Vinton was an early employee of the Consumer Financial Protection Bureau and an architect of IRS Direct File. She says state and local governments are still investing in digital teams, often with cost savings. Her core finding is blunt: there are no patterns. That has a bearing on how the startup sector should read its own disruption. The teams quietly making government services work better, she writes, have been or are being dismantled. What replaces them matters. For founders selling into government, the buyer's capacity is changing faster than any procurement cycle can track.
Supply chains and the cost of decoupling
On the hardware side, CNN reported on 3 October that US restrictions on Chinese technology are pushing robotics firms to rebuild supply chains at higher cost. Chef Robotics founder Rajat Bhageria told CNN that his robot arms, sold to food manufacturers, have relied heavily on Chinese components. Investors and customers now ask why he still uses foreign-made parts. The components in his arms are not yet affected by the bans, but he worries the list will grow. "You can imagine what's the next shoe to drop, right? Like, the next thing might be fixed robot arms," he said.
Ben Armstrong, executive director of MIT's Industrial Performance Center, told CNN that a lot of digital life is built on foreign-made components and goods assembled abroad. That will not change soon, he said. "We just don't know how to make them in the US," he said. "So, there will be a learning curve, and during that learning curve process, the price will be high." For startups, that is a cost story that arrives before any headcount story does.
Two quieter items also appear. Works in Progress argued on 4 October that genetic methods to suppress disease-carrying mosquitoes have spent fifteen years in US regulatory limbo. Field trials cut Aedes aegypti populations by 80 percent in the Cayman Islands and about 95 percent in Juazeiro, Brazil. The piece notes nearly 4,000 US dengue cases in 2024, a 360 percent increase over the previous decade's average of 830 a year. Startup Qatar's investment programme, dated 4 October on its site, offers up to $1.1 million for seed-stage companies and up to $5.5 million for growth-stage companies moving to Qatar, released against agreed milestones. Both are reminders that capital and regulation, not only layoffs, set the pace of restructuring.
One more post is worth naming because it captures the mood the layoff headlines feed on. Cory Dransfeldt's 4 October essay argues that users should reduce the value tech companies can extract from them. He suggests giving false birth dates, using throwaway email aliases, rejecting cookies and cancelling streaming subscriptions. It is an argument, not a dataset, and it does not mention layoffs. But the hostility it describes is the same current running under the restructuring coverage. Workers and users are being asked to absorb change while the companies doing the changing report the results on their own schedule.
The honest read is that the layoff story is real but under-sourced here, while the restructuring story is broader than headcounts. Neko Health is hiring into a New York clinic. Chick-fil-A is keeping humans on the drive-thru. Chef Robotics is hunting for non-Chinese parts. Government digital teams are gone, and state and local ones are still spending. None of that fits a single wave. The Knight paper's warning applies: methods, applications and adoption move on different clocks, and a news cycle is the fastest of the three.
Sources
10- 01America wants to wean itself off Chinese technology. Will the pain pay off?EN
- 02Spotify billionaire's body scan startup has come to AmericaEN
- 03Chick-fil-A rules out AI drive-thru ordering as fast-food rivals embrace technologyEN
- 04Your Car Is Sharing Data With Big Tech Companies, Study FindsEN
- 05What Did Civic Tech Learn in the Last Decade?EN
- 06AI as Normal Technology (2025)EN
- 07How many humans does it take to make tech seem human?EN
- 08Become worthless (to tech companies)EN
- 09The technology to eradicate mosquito-borne disease already existsEN
- 10Up to US$5.5M funding to move your startup to QatarEN
All figures and quotations in this text come from the sources listed below.
Content prepared by the editorial team with AI assistance.
Comments
0- No comments yet — be the first.