AI is leading to more layoffs and more flexible talent
AI cuts are not sticking: 55% of leaders who made AI layoffs call them the wrong call, 36% of companies that cut roles for AI have already rehired more than half of them, and much of that work is coming back as flexible talent.
By Matthew Mottola, Tony Buffum and Jeff Nugent · Published December 1, 2026
Key findings
55% of leaders who made AI layoffs call them the wrong call, and 36% of companies that cut roles for AI have already rehired more than half of them.
Source: Orgvue, 2025; Careerminds survey of HR leaders, February 2026; Robert Half, 2026; Resume.org, January 2026. Four surveys with different bases.
Exhibit 9
AI cuts are not sticking
Percent of each surveyed group
Leaders: AI cuts were the wrong call
55%
Rehired over half the roles cut
36%
Brought back roles cut for AI
32%
Rehiring cost more than cuts saved
31%
AI fully replaced a role
9%
Source: Orgvue, 2025; Careerminds survey of HR leaders, February 2026; Robert Half, 2026; Resume.org, January 2026. Four surveys with different bases.Source: 2026 Flexible Workforce Study, Human Cloud and Independently.
Challenger, Gray & Christmas counted 121,401 US job cuts attributed to AI through October 2026, about one in five of all announced cuts.
Source: Challenger, Gray & Christmas, October 2026 Job Cuts Report, November 5, 2026.
74% of US independents used generative AI in 2025, while only 36% of US workers say they have the training and resources to use AI at work.
Source: MBO Partners, State of Independence 2025; Jobs for the Future, survey of US workers, March 2026.
Exhibit 10
Independents are already applying AI while most workers wait on training
Percent of each group
Independents who use generative AI (MBO Partners, 2025)
74%
US workers who say they have the training and resources to use AI at work (Jobs for the Future, 2026)
36%
Source: MBO Partners, State of Independence 2025; Jobs for the Future, survey of US workers, March 2026.Source: 2026 Flexible Workforce Study, Human Cloud and Independently.
AI assistants and answer engines went from 9% of solution profile reads on Human Cloud in February 2026 to about 21% by October.
Source: Human Cloud platform data (profile-view events), pulled Nov 10, 2026. AI readers are assistants fetching a page at a person’s request and answer engines indexing profiles.
Exhibit 12
AI now makes about one in five profile reads
AI assistants and answer engines as a share of profile reads on Human Cloud, 2026
9%Feb
16%Apr
26%Jun
21%Aug
21%Oct
Source: Human Cloud platform data (profile-view events), pulled Nov 10, 2026. AI readers are assistants fetching a page at a person’s request and answer engines indexing profiles.Source: 2026 Flexible Workforce Study, Human Cloud and Independently.
A solution with a verified business case draws about 53 AI reads per profile, against 12 for a solution whose case is not yet verified and about 1 for a solution with none.
Source: Human Cloud platform data, pulled Nov 10, 2026.
Exhibit 13
AI reads proven solutions far more
AI reads per solution profile, Oct 12 to Nov 10, 2026
Solutions with a verified business case
53.1
Solutions with a business case, not yet verified
12.4
Solutions with no business case
1.1
Source: Human Cloud platform data, pulled Nov 10, 2026.Source: 2026 Flexible Workforce Study, Human Cloud and Independently.
7 in 10 talent platform business cases describe work the provider owned through to the outcome, against about 1 in 10 staffing cases, and solutions with a case describing work delivered draw about 1.7 times the AI reads.
Source: Human Cloud business cases, each read and sorted by hand, November 2026.
Exhibit 14
Talent platforms show work delivered, and AI reads it more
Share of buyer business cases on Human Cloud, by what the case describes
Talent platformsStaffing firms
The provider owned the work and the outcome
70%
9%
Placement only (hires, fill rate, time to fill)
28%
73%
Source: Human Cloud business cases, each read and sorted by hand, November 2026.Source: 2026 Flexible Workforce Study, Human Cloud and Independently.
The chapter in brief
Every employee is becoming a manager of agents. But AI activity is not yet AI value: only 13% of CIOs say their enterprise has realized significant or transformational value from AI so far.
The agents arrived alongside a wave of job cuts blamed on AI, and the cuts are not sticking. Much of that work returns as flexible work. That is good for speed and a risk for compliance, because a former employee doing the same job as a contractor is the textbook misclassification case.
An agent is only as good as the context it has. Somebody still has to decide what it should do, set it up inside a real business, check its work and fix it when it drifts. That is where flexible talent shines: specialized talent platforms already have the industry expertise, and their talent has the AI expertise.
AI is now part of how companies choose, and what it reads is not random. AI assistants favor solutions with verified business cases and solutions that describe work delivered rather than placements. The lesson is simple: publish named, verified outcomes, because that is what the agent passes along.
From the practitioners in this chapter
“At Cisco, we just rolled out in August: 90,000 employees have agents. So I have an agent, think about it as my digital worker.”
Bob Cicero
Global Director of Workplace Experience, Cisco
“We went really deep into all Git repositories. In open source since the beginning of Git, analyzed every line of code.”
Jose de Cabo
Founder, Remotely
Solutions named in this chapter
Each one has a profile on Human Cloud, with its verified business cases.
Ten chapters, 40 exhibits, business cases verified by named companies, and a one-page action plan for innovation, HR and procurement leaders and for the solutions that serve them.
