Skip to main content
Sercxi Index · Q2 2026 - Final

AI Engineering Displacement

EMEA · Q2 2026 · Final Assessment

Morgan Stanley puts the UK at 6% net AI-driven job losses - the worst reading among major developed economies. Germany shows 1% net gain. DeepL, Solaris, Bird, Renault, and Pleo all framed Q2 cuts as 'AI-native replatforming'. The EU AI Act has produced a net-new C-1 role no one has staffed.

AI-native replatforming is the layoff cover story of Q2.

Method

Final Q2 read grounded in Morgan Stanley AlphaWise, EU AI Act enforcement guidance, primary firm disclosures, and Christian & Timbers / ManpowerGroup / WEF talent data. Directional points explicitly flagged.

7 Roles Assessed·🟢 3 Stable🟡 2 Transitioning🟠 2 Exposed
🟢
StableRole intact, demand holding
🟡
TransitioningScope shifting materially
🟠
ExposedMandate erosion underway
🔴
DisplacedRole being eliminated

Key Findings

UK records 6% net AI-driven job losses over the prior 12 months - the worst reading among major developed economies (Morgan Stanley AlphaWise, May 2026).

DeepL -250 (~25%, May 2026), Solaris -20%, Bird -20%, Pleo same-day-as-agentic-launch cuts all framed as 'AI-native'.

EU AI Act GPAI obligations have been live since 2 Aug 2025; June 2026 amendments extend some HR-tech timelines to 2027/2028 but Morgan Lewis describes this as preparation time, not reprieve.

Head of AI Safety supply critically constrained - active mandates at Deutsche Bank, Faculty AI, J&J unfilled or in late stages.

Head of Agentic Systems is the fastest-emerging title with no established talent pipeline.

Meta Superintelligence Labs continues extracting EMEA research talent at compensation no European employer matches.

Methodology

The Sercxi Displacement Index assesses senior leadership roles against three structural vectors. Each is scored 1–5. The combined profile produces a Displacement Rating.

Elimination Risk(1–5)

The probability that the role is structurally removed from organisational charts within 24 months - not through attrition, but through deliberate elimination driven by automation, managed services, or mandate consolidation.

Redefinition Pressure(1–5)

The degree to which the role's scope, accountability, and required competencies are shifting. A high score indicates the job description is being rewritten faster than most incumbents are adapting.

Creation Signal(1–5)

The strength of net-new demand for the role or its evolved successor. High creation signals indicate structural tailwinds - new regulatory mandates, emerging technology domains, or market gaps creating durable hiring pressure.

Scorecard Overview

RoleEliminationRedefinitionCreationRating
Chief AI Officer
🟡Transitioning
VP Engineering (AI/ML)
🟠Exposed
Head of Applied Research
🟡Transitioning
Head of ML Platform
🟠Exposed
Head of AI Product
🟢Stable
Head of AI Safety / Responsible AI
🟢Stable
Head of Agentic Systems
🟢Stable

Role-by-Role Analysis

01

Chief AI Officer

Elimination: 2/5·Redefinition: 4/5·Creation: 3/5
🟡Transitioning

Demand is real but consolidating - boards want fewer, more empowered CAIOs rather than proliferating the title. London and Zurich lead placement activity; Paris is second-tier.

Many CAIOs at pre-AI-native firms face scope reduction as AI gets embedded into core engineering leadership. The seat survives where it owns measurable enterprise transformation; it shrinks where it remains advisory.

Directional only - no public benchmark on CAIO tenure across EMEA yet.

02

VP Engineering (AI/ML)

Elimination: 4/5·Redefinition: 4/5·Creation: 4/5
🟠Exposed

The primary zone of EMEA displacement. Companies replatforming to agentic stacks are eliminating legacy VP Eng layers or merging into product engineering.

Renault's June 2026 cuts (800 traditional engineering roles, AI recruitment expansion - caliber.az) exemplify the trade. Morgan Stanley's AlphaWise data put the UK at 6% net AI-driven job losses over the prior 12 months (resultsense.com, May 2026) - the worst reading in major developed economies.

Scarcity of candidates who can manage both inference infrastructure and product delivery keeps comp elevated in London and Amsterdam.

03

Head of Applied Research

Elimination: 2/5·Redefinition: 3/5·Creation: 3/5
🟡Transitioning

Supply extremely thin. PhDs with production deployment track records are routinely poached by US hyperscalers or GCC sovereign entities.

Meta's Superintelligence Labs (expanded Jun 2026 via Virtue AI acquisition - Axios) continues to pull talent out of European research units. Mistral (Paris) and DeepMind (London) partially offset but cannot compete on RSU value.

Sercxi sees the most active leakage from London labs; Paris is becoming an attractor city, particularly post-Brexit visa changes.

04

Head of ML Platform

Elimination: 3/5·Redefinition: 4/5·Creation: 4/5
🟠Exposed

The hottest operational hire across EMEA FSI in Q2. Deutsche Bank, BNP, UBS, and ING are all in market.

Agentic stack transitions require GPU-cluster orchestration, inference serving, and RAG pipeline expertise simultaneously - a profile that did not exist as a single seat 18 months ago.

DeepL's restructure (~250 roles, May 2026 - aiineurope.co) signals mid-market AI-native firms eliminating older ML-ops layers and rehiring at platform-lead level.

05

Head of AI Product

Elimination: 2/5·Redefinition: 4/5·Creation: 4/5
🟢Stable

A growing role; the candidate pool is widening faster than demand. EU AI Act conformity assessments and transparency disclosures elevate product leaders who understand regulatory interfaces.

Amsterdam and Berlin are strongest; London demand softened slightly in the 6% net-loss environment.

The brief is moving from feature roadmap ownership to regulatory product packaging.

06

Head of AI Safety / Responsible AI

Elimination: 1/5·Redefinition: 1/5·Creation: 5/5
🟢Stable

A structural must-have post-GPAI enforcement (live since 2 Aug 2025; updated guidance Mar 2026 - artificialintelligenceact.eu). Active live vacancies: Deutsche Bank Director - Safe AI (London, Jun 2026), Faculty AI Technical Director AI Safety (London, Mar 2026), J&J Sr Director AI Safety (Beerse/Zug).

Supply is critically constrained - directionally, fewer than ~400 practitioners globally with both technical depth and EU governance literacy.

Our directional read is a 30-45% premium over equivalent Head of ML Platform rates in London.

07

Head of Agentic Systems

Elimination: 1/5·Redefinition: 1/5·Creation: 5/5
🟢Stable

The fastest-emerging title in EMEA. Pleo, Klarna, Revolut, and N26 are each building or expanding agentic product capability. Pleo conducted layoffs the day after launching its agentic finance suite (thenextweb.com, Jun 2026) - the clearest architectural-shift signal of the quarter.

No established talent pipeline. Most hires are lateral conversions from senior ML platform or AI product backgrounds.

Compensation benchmarks for EMEA are not yet published.

The Sercxi Read

EMEA AI engineering is bifurcating sharply - not between winners and losers, but between roles being eliminated at speed and roles where supply is so constrained that even well-capitalised employers cannot fill. The former category (traditional VP Eng, ML Ops leads, AI Product generalists) is seeing significant churn behind the 'AI-native transformation' vocabulary that has become Q2's dominant HR language.

EU AI Act enforcement has produced a talent bottleneck at the governance layer. Head of AI Safety and Head of Responsible AI are being opened simultaneously at FSI institutions, model providers, and the new wave of GPAI deployers in legal, health, and insurance. There are not enough qualified candidates. Executives who combine deep ML systems experience with working knowledge of Articles 51-55 obligations command a premium no published benchmark yet captures - directionally 30-45% over equivalent Head of ML Platform rates in London.

The wildcard is Paris. Mistral's continued growth and sustained French government AI investment (post-Renault) position the city as an attractor for mid-senior talent displaced from London. We see this in our own mandates but the data to quantify it does not yet exist publicly.

'AI-native' is the word boards use when they cannot say 'overstaffed'.

Your Three Questions

Answer these honestly. No form. No follow-up unless you want one.

1.

As 'AI-native' restructurings multiply, at what point does UK employment tribunal caseload force companies to evidence genuine role elimination - and how does that change the speed of AI engineering headcount decisions?

2.

Which EMEA model provider (Mistral, Aleph Alpha, DeepMind London) will make the first credible counter-offer to Meta Superintelligence Labs compensation - and through what instrument?

3.

Is the GPAI enforcement talent gap severe enough to warrant a dedicated EMEA AI governance talent firm, and who is best positioned to build it?

Each question is a vacancy waiting for a credible candidate.

Initiate Confidential Briefing →

Save this report

Print-optimized layout for executive distribution.

Q3 2026 · September 2026

Q3 2026 - Post-AI Act Read

Post-2 Aug 2026 enforcement read, UK tribunal-caseload analysis on 'AI-native' redundancies, and updated agentic-systems mapping.