They told you AI would take your job. Then they changed their story.
The forecast softened. The filings did not.
A year ago the loudest voices in technology said AI would wipe out white-collar work. Last month, almost in unison, they took it back: growth, partnership, people at the centre of everything. In the same month, AI was the leading cited reason for US job cuts for the fourth month running, and the analysts who watch this closely predict half of those cuts will be quietly rehired -- offshore, at significantly lower salaries. The story changed because it was bad business. The substitution did not. This issue reads the walk-back from the receiving end: Mumbai, Dubai and Nairobi, where the rerouted work will land, repriced.
CEOs still forecasting significant AI job cuts. Down from 46% in January 2025.
Fifty-five per cent of employers already regret an AI layoff. The walk-back is not an apology, it is narrative management, and the numbers underneath it did not walk back: record AI-attributed cuts, three in four customer-facing deployments rolled back, and the quiet rehiring routed offshore at a discount. Trust the filings, not the interviews.
- Track filings, not interviews because the annual report and the monthly cut count are where substitution shows up under legal liability; the conference stage is where it is managed.
- Price the rollback before the saving and require every AI substitution case in your organisation to carry a reinstatement cost line, because three in four customer-facing deployments have already been pulled back.
- Position for the rebuilt rung by naming which function your industry cut onshore this year and deciding whether you will run its offshore rebuild or compete against it.
Rising
- AI deployment auditor↗
- Escalation architect↗
- AI training-data engineer↗
Cuts this wk
- Microsoft−4,800
CEOs still forecasting significant AI job cuts. Down from 46% in January 2025.
Executive doom halved in sixteen months, and nothing in the data moved with it: Challenger logged AI as the leading cited reason for US job cuts for the fourth consecutive month in June at 14,029, taking the half-year total to 101,743, about 23% of every announced cut, while Microsoft, Meta and Oracle kept filing reductions the interviews no longer dwell on.
The wipeout was a pitch. The reassurance is one too.
A year ago the wipeout story primed investors for margin expansion; now the same companies need employees to adopt the tools they were told would replace them, and a workforce where only 6% expect AI to bring more opportunity does not adopt willingly -- so the story flipped to partnership, and the only line that did not move is the attribution row in the layoff filings.
Fifty-five per cent of employers already regret an AI layoff. The walk-back is not an apology, it is narrative management, and the numbers underneath it did not walk back: record AI-attributed cuts, three in four customer-facing deployments rolled back, and the quiet rehiring routed offshore at a discount. Trust the filings, not the interviews.
- Track filings, not interviews because the annual report and the monthly cut count are where substitution shows up under legal liability; the conference stage is where it is managed.
- Price the rollback before the saving and require every AI substitution case in your organisation to carry a reinstatement cost line, because three in four customer-facing deployments have already been pulled back.
- Position for the rebuilt rung by naming which function your industry cut onshore this year and deciding whether you will run its offshore rebuild or compete against it.
Career vectors.
Two weeks of named layoffs. 6 rising role categories with sourced hiring signals.
Announced layoffs · week-on-week
Rising role categories
Hiring signal · named companies · this week
AI deployment auditor
Sinch found 74% of enterprises rolled back or shut down deployed AI customer agents. Someone has to establish what failed, what it cost and what returns.
Escalation architect
Commonwealth Bank of Australia reinstated more than 40 service roles after its voice bot failed the call queue. Designing the human handoff is now a job.
AI training-data engineer
Ford is rehiring experienced engineers because the tool is only as good as the information used to train it. Data quality is the binding constraint.
Senior AI specialist
Naukri logged AI hiring up 16% while overall IT listings fell 3%, with demand shifting to senior and specialised talent. The middle of the ladder is the market.
AI skills translator
PwC counts 77 new skills demanded in the UAE's most AI-exposed roles against 22 in the least. Sequencing that curriculum inside an organisation is a role.
Rebuild programme lead
Forrester predicts half of AI-attributed layoffs will be quietly rehired offshore at lower salaries. The reroute needs people who can run it from the receiving end.
Three regions. Three speeds.
Short read · this week's signal through the India, Middle East, and Africa lens
India is running the walk-back's two halves at once. Naukri's June data shows the split-screen: AI hiring up 16% while overall IT listings fell 3%, demand migrating to senior and specialised talent exactly as the zero-to-two-year rung thins. And India is the named destination of the regret cycle: when Forrester says half of AI-attributed layoffs return quietly offshore at significantly lower salaries, the capability centres of Bengaluru and Hyderabad are where 'offshore' mostly means. The work comes back. The question is at what price, and who supervises the machine it now sits beside.
The UAE's AI labour market tripled its share of job postings in four years, and PwC's barometer puts a number on what the walk-back obscures: the most AI-exposed roles now demand 77 new skills against 22 in the least exposed. The Gulf's bet is that it can hire and train through the turbulence rather than cut through it, and its policy levers, from Emiratisation quotas to DIFC and ADGM licensing, give it a means of steering who fills the new roles. Reassurance rhetoric changes nothing about that skills arithmetic.
If half the AI-attributed cuts are quietly rehired offshore at lower salaries, the reroute does not stop at India. Kenya's four largest BPO operators formed an alliance in February targeting 100,000 additional jobs on a base of 60,000, and the discounted second chance the regret cycle creates is exactly the volume they are positioned to absorb. Africa arrives as the rung is rebuilt, cheaper, and the continent's task is to take the volume without inheriting the fragility: rungs rebuilt on price alone can be cut again by the same logic that cut them the first time.
Nine sectors. Nine weathers.
Short read · this week's signal across the nine sectors we cover
The narrative flip is this sector's own product launch: Microsoft, Meta and Oracle kept cutting while their peers' rhetoric softened into growth language.
The regret cycle is billable: rollback remediation, AI ROI audits and rebuild programmes are new revenue lines, and the evidence cuts both ways, with heavy AI adopters holding employment 10.2% higher than non-adopters.
Commonwealth Bank of Australia's reinstated service roles are the first big regulated-industry rollback, and every bank's automation case now gets read against it.
Customer-service automation is the epicentre of the 74% rollback figure, and retail ran the largest share of those deployments.
Slower adoption just became an advantage: the sector gets the rollback lesson from retail and banking without paying the tuition.
Ford's rehiring of experienced engineers is the sector's tell: industrial AI fails on training data before it fails on models.
Front-desk and guest-service automation is queued to repeat retail's rollback arc, one budget cycle behind.
The reroute adds seats offshore: capability-centre absorption in Indian and African hubs is the space demand created by the regret cycle.
AI-attributed cuts have barely touched the sector this half; its exposure remains demand-side, powering the compute the narrative war is fought over.
Five skills to master this week.
For Editor reAImagine · curated to this issue's signal · 90-day horizon
Rollbacks at three in four deployments create a paid discipline, and the QA lead becomes the AI deployment auditor.
The gap between the pitch and the filing is measurable, and the finance analyst becomes the AI ROI auditor.
Banks are rehiring the humans their bots replaced, and the service manager becomes the escalation architect.
Coding alone is no longer the job, and the mid-career developer becomes the domain product engineer.
Half the AI-attributed cuts are predicted to return offshore at lower cost, and the delivery manager becomes the rebuild programme lead.
Issue 015 opened Ledger 001 with five dated forecasts and a promise: scored from Issue 016 onwards, wrong calls stay on the record. This is the first scoring, and it opens with a miss. The claim below is reproduced exactly as published; the score and the reasoning sit beneath it, at full length, because a ledger that mumbles its misses is a marketing page. Four entries remain open and carry forward with their resolve dates. Ledger 002 opens below them, written in the same discipline it will be judged by.
- LEDGER 001 · FIRST SCORING
- LEDGER-001-02✗ MISS
Anthropic's ID-verification policy takes effect and, whatever its stated intent, functions in practice as a citizenship-sorted access path: US consumers regain restricted-tier access first, with no announced parity path for Indian or GCC passport holders. Anthropic says the change is an unrelated appeals update; we forecast the observable outcome and will score it.
Scored 9 July 2026. Fable 5 came back for every consumer on earth on the same day, 1 July, because the US Commerce Department lifted the export controls on 30 June. The restoration ran through diplomacy, not identity checks, and it landed a week before the ID policy took effect on 8 July. The policy itself verifies identity and age for flagged consumer accounts, carries no nationality component at all, and exempts Team, Enterprise and API customers. The disconfirming evidence we carried inside the entry, Anthropic's statement that this was an unrelated appeals update, held up better than our forecast did. The one US-first pattern that did appear, Mythos 5 returning on 26 June for a government-approved list of US organisations, ran through a different mechanism than the one we named, and a forecast does not get credit for a different mechanism. To score this a hit we needed restricted access re-sorted by passport through the verification flow. It was not.
- LEDGER-001-01OPEN
At least one further US frontier-model release goes through government pre-release review rather than open launch, extending the pattern already visible in June.
- LEDGER-001-03OPEN
At least one of TCS, Infosys, Wipro or HCLTech publicly announces a formal multi-model or sovereign-fallback architecture policy as strategy, not as a procurement footnote.
- LEDGER-001-04OPEN
The first senior role explicitly titled for AI sovereignty or model continuity, distinct from CISO or Chief AI Officer, is publicly posted by a GCC entity or Gulf sovereign-linked employer.
- LEDGER-001-05OPEN
Accenture's new bookings decline year on year again, confirming the June repricing as structural rather than sentiment.
- LEDGER 002 · NEW THIS ISSUE
- LEDGER-002-01
At least one company that attributed 2026 layoffs to AI is publicly reported to have rebuilt the same function in India, the Gulf or Africa, directly or through a capability centre or outsourcing partner, within twelve months of the cut.
- LEDGER-002-02
Challenger's AI-attributed US job-cut count for the second half of 2026 exceeds the first half's 101,743, despite the softened executive rhetoric. The narrative and the number diverge further, not less.
- LEDGER-002-03
India's top four IT services firms, TCS, Infosys, Wipro and HCLTech, in aggregate add net headcount over FY27's first half, April to September 2026, while each scales AI-attributed revenue, confirming the reroute: the work returns offshore even as the Western rhetoric softens.