AI at work: jobs, rules and workplace control

The single biggest theme this week is plain and uncomfortable: AI is no longer just a productivity story for HR — it is an employment story. From headline headcount moves and warnings from economists to a burst of new tools that can act on behalf of employers, the evidence is converging that organisations must manage not only adoption but displacement, compliance and workplace control.

The jobs test: cuts, warnings and the shape of displacement

Redundancies tied to AI are moving from theoretical to operational. LHH's survey makes that explicit: 87% of UK HR leaders report they have made or plan redundancies linked to AI transformation, and the report flags gaps in redeployment support that HR must close LHH: 87% of UK HR plan AI-driven redundancies. That corporate reality sits alongside warnings from economists and researchers: nearly 200 AI researchers and more than a dozen Nobel economists signalled that transformative AI risks large-scale job disruption and broader economic upheaval Nobel economists and AI researchers warn of job disruption. Real‑world layoffs are already landing — the Xbox restructuring at Microsoft removed thousands and prompted union pushback, a reminder that cuts tied to strategy shifts are sometimes met with organised resistance Xbox layoffs: Microsoft restructures, unions push back.

Collectively these items are a call to action for HR: plan for replacement scenarios and for redeployment pathways, update consultation and change‑management processes, and test your internal safety nets for people otherwise set to be displaced.

Regulation and enforcement: a tighter perimeter around hiring and monitoring

Regulators are not letting employers set the rules alone. In the US the EEOC has signalled an enforcement focus on AI in hiring with its draft FY2026–2030 plan and related rule changes from federal and city agencies that demand disclosures, bias audits and stronger enforcement EEOC Draft Plan Flags AI Risk in Hiring; the EEOC, New York City and several states have tightened AI hiring rules and disclosure requirements EEOC, NYC and States Tighten AI Hiring Rules. New York’s Local Law 144 remains a live compliance pressure point for employers operating in the city Local Law 144: NYC employers face rising compliance pressure.

In the UK and EU the pivot is toward governance and worker protections: the Commons Business and Trade Committee has been taking evidence on jobs, redundancy risks and skills — material HR should feed into policy responses and consultations Commons probes AI and the future of the workforce. The UK government’s 'Make Work Pay' consultation opens a debate about workplace monitoring and privacy that will affect HR analytics and surveillance policies Workplace monitoring: 'Make Work Pay' consultation. Across the channel, the European Commission is moving on cybersecurity for advanced AI and voluntary transparency codes for AI-generated content — changes that will reshape procurement, testing and vendor expectations EU action plan: advanced AI cybersecurity and EU Code for AI-generated content transparency. The Digital Omnibus has also amended timing and enforcement for the AI Act, forcing vendors and buyers to rework roadmaps Digital Omnibus changes EU AI Act compliance.

New tools, new workplace levers — and new risks

This week saw a flurry of product moves that make the technical risk immediate. OpenAI released GPT‑5.6 and ChatGPT Work — enterprise‑grade agents that can access apps and files and therefore create direct data‑protection and operational risk for employers OpenAI launches GPT-5.6 and ChatGPT Work. WorkJam announced an autonomous layer to assign, prioritise and verify frontline work with auditable logs and manager override — a concrete example of automation moving to execution at scale WorkJam adds autonomous AI layer for frontline work. iCIMS added analytics to high‑volume hiring to accelerate screening iCIMS adds analytics to high-volume hiring, while Microsoft is offering a practical control: an in‑meeting toggle to disable Teams’ meeting AI features — a small technical control with big HR implications for consent and privacy at meetings Microsoft adds in‑meeting toggle for Teams Meeting AI.

These products change the locus of control: decisions can be made by agents and models, not just managers. That raises questions about audit trails, appeal routes, and who owns the outcomes when an autonomous system prioritises work or rejects a candidate.

Skills, governance and the vendor workforce: who builds and who watches

As adoption accelerates, vendors and employers are also trying to build the skills and governance layers. Cognizant’s pledge to train and certify 5,000 Frontier engineers and 10,000 Frontier operators is a reminder that new technical roles — and the governance they embody — are being staffed by both vendors and clients Cognizant vows 5,000 Frontier engineers, 10,000 operators. Gov.UK’s Cloud Challenge Book stresses building cloud and AI capability across the public sector, signalling that skilling and secure procurement will be central to public‑sector HR plans Cloud Challenge Book 2026: Gov.UK on cloud and AI skills. Workday’s expansion through HR Path in the UK & Ireland shows how HR platforms are doubling down on local implementation capacity and the practical work of embedding AI into HR systems Workday expands EMEA VAR with HR Path UK&I.

HR teams should be explicit about who will set model standards, approve vendor updates, run bias audits and hold the remediation budget when a tool misfires.

UK vs US: tempo and exposure

The UK story this week is about the immediate operational impact and a policymaking conversation: high proportions of HR planning AI‑linked redundancies LHH: 87% of UK HR plan AI-driven redundancies, active parliamentary scrutiny Commons probes AI and the future of the workforce and a government consultation on monitoring Workplace monitoring: 'Make Work Pay' consultation. The US picture is more enforcement‑heavy and vendor‑driven: regulatory agencies are tightening hiring rules and signalling enforcement, city laws such as Local Law 144 are being enforced more aggressively Local Law 144: NYC employers face rising compliance pressure, and major vendors are shipping agentic products that elevate data and operational risk OpenAI launches GPT-5.6 and ChatGPT Work. In short: the UK is wrestling with displacement and the social compact; the US is confronting compliance and vendor risk at scale.

What to watch

Next week, HR leaders should watch the EEOC’s comment window and agency signals closely: the draft FY2026–2030 plan is open for comment through July 19 and will indicate the urgency and direction of enforcement around hiring tools EEOC Draft Plan Flags AI Risk in Hiring. Expect employers to be asked for sharper disclosures, stronger bias‑testing protocols and clearer remediation plans.

On the regulatory side in Europe and the UK, keep an eye on how the Digital Omnibus changes are implemented and how the Article 50 transparency requirements are translated into vendor contracts; the EU’s push on advanced AI for cybersecurity could also change procurement criteria for HR tech vendors Digital Omnibus changes EU AI Act compliance EU action plan: advanced AI cybersecurity. Finally, watch vendor governance moves: product updates from OpenAI, Microsoft and frontline automation vendors will materially affect privacy, audit trails and redeployment planning inside organisations OpenAI launches GPT-5.6 and ChatGPT Work Microsoft adds in‑meeting toggle for Teams Meeting AI.