[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fptbzlgpCRjnGVs7Y3p3Zm9cGNtAvJt-Eiwri9cCGJQg":3,"$flTuTpQtxo9WdQU3ywWE0B3HcD3_m4wRNgN123QthFwQ":29},{"success":4,"data":5},true,{"id":6,"slug":7,"title":8,"excerpt":9,"content":10,"category":11,"tags":12,"author":19,"cover_image_url":20,"reading_time_minutes":21,"is_published":4,"published_at":22,"created_at":23,"updated_at":24,"author_avatar":20,"is_featured":25,"meta_title":26,"meta_description":27,"meta_keywords":28},"49a008f0-3a61-49c6-8957-e2381395d437","ai-hiring-correction-ford-ibm-2026","Ford Rehired 350 Engineers. IBM Is Tripling Entry-Level Hiring. What the 2026 AI Correction Actually Teaches Recruiting Teams","Ford rehired 350 engineers. IBM automated 94% of HR requests and is tripling entry-level hiring anyway. What the 2026 AI correction actually teaches recruiting teams about where the human belongs in a workflow.","\u003Cp>For the past two years, the dominant hiring story was \"AI is replacing entry-level and mid-career work.\" In 2026, a second story started running alongside it: some of the same companies that cut for AI are now paying, in cash and in reputation, to bring humans back. Not because the AI stopped working — because it was doing a different job than the one the org chart assumed.\u003C\u002Fp>\n\n\u003Cp>Three cases, all independently reported this year, make the pattern visible. None of them is a story about AI failing outright. Each is a story about where inside a workflow AI was asked to make the call versus where a human still had to.\u003C\u002Fp>\n\n\u003Ch2>Ford: 350 engineers, hired back to fix what automation missed\u003C\u002Fh2>\n\n\u003Cp>Ford has spent the last three years rehiring experienced engineers — 350 of them, many former employees or people pulled from suppliers — specifically to work on the quality problems its automated systems weren't catching. Ford's VP of vehicle hardware engineering, Charles Poon, told reporters the company had mistakenly believed it could swap in AI-driven processes and still hold product quality, according to reporting from \u003Ca href=\"https:\u002F\u002Ftechcrunch.com\u002F2026\u002F06\u002F28\u002Fford-rehires-gray-beard-engineers-after-ai-falls-short\u002F\" rel=\"noopener\">TechCrunch\u003C\u002Fa> and \u003Ca href=\"https:\u002F\u002Fwww.bloomberg.com\u002Fnews\u002Farticles\u002F2026-06-25\u002Fford-has-been-rehiring-quality-inspectors-after-ai-fell-short\" rel=\"noopener\">Bloomberg\u003C\u002Fa> in late June 2026. The returning engineers weren't brought back to do the job AI was doing — they were brought back to mentor junior staff, rebuild the data pipelines feeding Ford's AI training, and fix the automated systems they had originally been slated to replace.\u003C\u002Fp>\n\n\u003Cp>The result showed up in the numbers: Ford topped JD Power's 2026 initial quality study, which tracks problems owners report in a vehicle's first 90 days. CEO Jim Farley told Bloomberg TV the veteran-engineer push helped cut warranty and recall costs by \"literally hundreds and hundreds of millions of dollars\" in Ford's favor. \u003Ca href=\"https:\u002F\u002Fwww.cnbc.com\u002F2026\u002F07\u002F01\u002Femployers-who-laid-off-workers-for-ai-are-reversing-their-decisions.html\" rel=\"noopener\">CNBC's July 2026 roundup\u003C\u002Fa> of employers reversing AI-driven cuts places Ford alongside a wider trend, not as an isolated story.\u003C\u002Fp>\n\n\u003Cp>What Ford's account actually describes is a gap between AI handling routine quality inspection and a residual class of problems — the ones that show up as recalls, not as line-item defects — that needed the judgment of someone who has seen a production run go wrong before. AI wasn't removed from the process. Humans were put back at the point where the automated system's misses became expensive.\u003C\u002Fp>\n\n\u003Ch2>IBM: AI resolves 94% of HR requests. It's still tripling entry-level hiring.\u003C\u002Fh2>\n\n\u003Cp>IBM's case is more direct, because IBM built the AI in question and is talking about it in public. AskHR, IBM's internal HR assistant, now resolves roughly 94% of routine employee requests. The remaining 6% — the requests involving ethical judgment calls and genuine exceptions — still land with human HR staff.\u003C\u002Fp>\n\n\u003Cp>Announcing this at Charter's Leading with AI Summit in February 2026, IBM's chief human resources officer Nickle LaMoreaux said the company would triple its U.S. entry-level hiring in 2026, \"across the board\" rather than in one department, according to \u003Ca href=\"https:\u002F\u002Fwww.bloomberg.com\u002Fnews\u002Farticles\u002F2026-02-12\u002Fibm-plans-to-triple-entry-level-hiring-in-the-us-in-2026\" rel=\"noopener\">Bloomberg's\u003C\u002Fa> and \u003Ca href=\"https:\u002F\u002Ftechcrunch.com\u002F2026\u002F02\u002F12\u002Fibm-will-hire-your-entry-level-talent-in-the-age-of-ai\u002F\" rel=\"noopener\">TechCrunch's\u003C\u002Fa> reporting from the summit — a decision that stood out precisely because it ran against the wave of tech layoffs attributed to AI efficiency gains that same year.\u003C\u002Fp>\n\n\u003Cp>The detail worth sitting with is what LaMoreaux did to the job descriptions themselves, not just the headcount plan. She rewrote entry-level roles to de-emphasize the tasks AI already handles well — routine coding, routine ticket resolution — and re-weight them toward the work that starts where AI's competence ends: engaging directly with customers, correcting AI outputs, communicating judgment calls to managers. That is a recruiting decision, not just an HR-automation decision. IBM didn't just decide humans were still needed; it decided \u003Cem>which part\u003C\u002Fem> of the job needed to be a human's part, and hired against that redefinition.\u003C\u002Fp>\n\n\u003Ch2>Commonwealth Bank: the correction isn't a full retreat\u003C\u002Fh2>\n\n\u003Cp>Not every case resolves as cleanly, and the honest version of this story includes the one that didn't. In 2025, Commonwealth Bank of Australia reversed a plan to cut 45 contact-centre roles after deploying an AI voice-bot meant to reduce call volume — the bot increased the workload on remaining staff instead of cutting it, and the bank ended up offering overtime and pulling team leaders onto phones to cope, according to reporting from \u003Ca href=\"https:\u002F\u002Fwww.finextra.com\u002Fnewsarticle\u002F46482\u002Fcommbank-reverses-plan-to-replace-call-centre-staff-with-ai\" rel=\"noopener\">Finextra\u003C\u002Fa> and \u003Ca href=\"https:\u002F\u002Fdig.watch\u002Fupdates\u002Fcba-reverses-ai-driven-job-cuts-after-union-pressure\" rel=\"noopener\">Digital Watch Observatory\u003C\u002Fa>. The Finance Sector Union took the case to Australia's Workplace Relations Tribunal, and CBA acknowledged its redundancy assessment hadn't properly accounted for actual call volumes.\u003C\u002Fp>\n\n\u003Cp>That reversal did not mark a general retreat from AI-driven headcount decisions at CBA. The bank subsequently cut further contact-centre and contractor roles as it continued expanding AI in customer service, per \u003Ca href=\"https:\u002F\u002Fwww.hcamag.com\u002Fau\u002Fspecialisation\u002Fhr-technology\u002Fcba-axes-contractor-call-centre-jobs-after-ai-rollout\u002F584060\" rel=\"noopener\">HR Director\u002FHCAmag\u003C\u002Fa> and \u003Ca href=\"https:\u002F\u002Fia.acs.org.au\u002Farticle\u002F2026\u002Fai-drives-fresh-commbank-job-cuts.html\" rel=\"noopener\">Information Age\u002FACS\u003C\u002Fa> reporting through 2026. The lesson from CBA isn't \"AI failed, humans won.\" It's narrower and more useful: the specific redundancy decision was made on a volume assumption nobody had actually measured, and got corrected once a union forced the measurement. Elsewhere in the same organization, AI-driven cuts that were measured more carefully went ahead.\u003C\u002Fp>\n\n\u003Ch2>The pattern underneath all three\u003C\u002Fh2>\n\n\u003Cp>Line these up and the common thread isn't \"AI can't do the job.\" It's that each org initially drew the human\u002FAI line at \u003Cem>task completion\u003C\u002Fem> — can the model do what the average request looks like — instead of at \u003Cem>exception handling and judgment\u003C\u002Fem>. Ford's AI could run a production line; it couldn't yet catch the failure mode that turns into a recall. IBM's AskHR resolves the routine 94%; the 6% that involves an ethical call still needs a person, and IBM is now hiring and training for exactly that 6%. CBA's bot could technically field calls; nobody had checked whether it actually reduced the total work, and the gap showed up as unpaid overtime until a union made someone check.\u003C\u002Fp>\n\n\u003Cp>That is not an argument against using AI in a hiring or workforce pipeline. IBM's own numbers make the opposite case: automating 94% of a workload and redeploying the freed capacity toward judgment calls is a defensible, even aggressive, AI strategy — it just isn't the same strategy as \"automate the workflow and cut the headcount that used to run it.\" The distinction is where the human sits, not whether one exists.\u003C\u002Fp>\n\n\u003Ch2>What this means for how you build a hiring pipeline\u003C\u002Fh2>\n\n\u003Cp>The same logic applies one level up, to the hiring process itself. A resume-screening tool or an AI interview can legitimately absorb the volume that used to overwhelm a recruiting team's calendar — that's the 94% case. What it shouldn't absorb, on the evidence above, is the decision that determines whether someone gets an offer. That decision is the organization's version of IBM's 6%: it's where judgment about fit, context, and exceptions actually lives, and it's exactly the layer that gets expensive to have gotten wrong, the way Ford's recalls and CBA's union case were expensive.\u003C\u002Fp>\n\n\u003Cp>Concretely, for a recruiting team reading the Ford\u002FIBM\u002FCBA pattern:\u003C\u002Fp>\n\n\u003Col>\n  \u003Cli>\u003Cstrong>Let AI do triage, not verdicts.\u003C\u002Fstrong> A screening stage should output scores, strengths, gaps and a recommendation for a human to act on — not a pass\u002Ffail that a candidate never sees a person weigh in on. NiceHire's own AI Screening Interview stage works this way: a completed interview stores structured technical, communication and cultural-fit scores, an overall recommendation, and recorded strengths and gaps against the application, feeding into the Technical Assessment, Live Interview and Final Round stages that follow it — it produces a recommendation for the pipeline, not a final decision.\u003C\u002Fli>\n  \u003Cli>\u003Cstrong>Measure before you cut, not after a complaint forces you to.\u003C\u002Fstrong> CBA's error wasn't using an AI voice-bot — it was skipping the step of checking whether the bot actually reduced total call volume before redundancies were finalized. Any AI tool that's meant to reduce a hiring team's workload deserves the same before\u002Fafter measurement, not an assumption.\u003C\u002Fli>\n  \u003Cli>\u003Cstrong>Rewrite the job, not just the headcount, around where judgment lives.\u003C\u002Fstrong> LaMoreaux's move at IBM wasn't just \"hire more people\" — it was redesigning entry-level roles around the work AI can't yet do well. The same applies to writing job descriptions for AI-adjacent recruiting or support roles: describe the exception-handling and judgment work, because that's the part of the job that isn't going anywhere.\u003C\u002Fli>\n\u003C\u002Fol>\n\n\u003Cp>The 2026 correction isn't AI in retreat. It's the year several large employers found out, the expensive way, exactly which slice of a workflow was never going to be a task a model could finish alone — and started hiring, and designing job descriptions, around that slice instead of ignoring it.\u003C\u002Fp>\n\n\u003Chr>\n\n\u003Cp>\u003Cem>Sources, all accessed 6 August 2026 via web search of the outlets' own reporting: \u003Ca href=\"https:\u002F\u002Ftechcrunch.com\u002F2026\u002F06\u002F28\u002Fford-rehires-gray-beard-engineers-after-ai-falls-short\u002F\" rel=\"noopener\">TechCrunch — \"Ford rehires 'gray beard' engineers after AI falls short\"\u003C\u002Fa> (28 June 2026); \u003Ca href=\"https:\u002F\u002Fwww.bloomberg.com\u002Fnews\u002Farticles\u002F2026-06-25\u002Fford-has-been-rehiring-quality-inspectors-after-ai-fell-short\" rel=\"noopener\">Bloomberg — \"Ford Has Been Rehiring Quality Inspectors After AI Fell Short\"\u003C\u002Fa> (25 June 2026); \u003Ca href=\"https:\u002F\u002Fwww.cnbc.com\u002F2026\u002F07\u002F01\u002Femployers-who-laid-off-workers-for-ai-are-reversing-their-decisions.html\" rel=\"noopener\">CNBC — \"Employers who laid off workers citing AI are already starting to regret it\"\u003C\u002Fa> (1 July 2026); \u003Ca href=\"https:\u002F\u002Fwww.bloomberg.com\u002Fnews\u002Farticles\u002F2026-02-12\u002Fibm-plans-to-triple-entry-level-hiring-in-the-us-in-2026\" rel=\"noopener\">Bloomberg — \"IBM Plans to Triple Entry-Level Hiring in the US in 2026\"\u003C\u002Fa> (12 February 2026); \u003Ca href=\"https:\u002F\u002Ftechcrunch.com\u002F2026\u002F02\u002F12\u002Fibm-will-hire-your-entry-level-talent-in-the-age-of-ai\u002F\" rel=\"noopener\">TechCrunch — \"IBM will hire your entry-level talent in the age of AI\"\u003C\u002Fa> (12 February 2026); \u003Ca href=\"https:\u002F\u002Fwww.finextra.com\u002Fnewsarticle\u002F46482\u002Fcommbank-reverses-plan-to-replace-call-centre-staff-with-ai\" rel=\"noopener\">Finextra — \"CommBank reverses plan to replace call centre staff with AI\"\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fdig.watch\u002Fupdates\u002Fcba-reverses-ai-driven-job-cuts-after-union-pressure\" rel=\"noopener\">Digital Watch Observatory — \"CBA reverses AI-driven job cuts after union pressure\"\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fwww.hcamag.com\u002Fau\u002Fspecialisation\u002Fhr-technology\u002Fcba-axes-contractor-call-centre-jobs-after-ai-rollout\u002F584060\" rel=\"noopener\">Human Resources Director \u002F HCAmag — \"Commonwealth Bank used AI to cut hundreds of offshore customer service jobs\"\u003C\u002Fa>; \u003Ca href=\"https:\u002F\u002Fia.acs.org.au\u002Farticle\u002F2026\u002Fai-drives-fresh-commbank-job-cuts.html\" rel=\"noopener\">Information Age \u002F ACS — \"AI drives fresh CommBank job cuts\"\u003C\u002Fa>.\u003C\u002Fem>\u003C\u002Fp>","Industry Trends",[13,14,15,16,17,18],"AI correction","Ford","IBM","workforce planning","AI in HR","hiring","NiceHire Team",null,9,"2026-08-14T01:35:17.305+00:00","2026-08-14T01:35:17.283918+00:00","2026-08-14T01:35:17.379911+00:00",false,"The 2026 AI Hiring Correction: Ford, IBM and CBA Lessons","Ford rehired 350 engineers, IBM is tripling entry-level hiring while automating 94% of HR requests. What the 2026 AI correction teaches recruiting teams.","AI correction, Ford, IBM, workforce planning, AI in HR, hiring",{"success":4,"data":30},{"posts":31,"count":34,"hasMore":25},[32],{"id":6,"slug":7,"title":8,"excerpt":9,"category":11,"tags":33,"author":19,"cover_image_url":20,"reading_time_minutes":21,"published_at":22},[13,14,15,16,17,18],1]