Skip to main content
Romania IT Market

Bosch is cutting 510 jobs in Timișoara. What it says about AI, and what it doesn't

Calin Muresan•
#Bosch#Timișoara#shared services#AI and jobs#engineering hiring

Bosch is cutting 510 jobs in Timișoara. What it says about AI, and what it doesn’t

By Calin Muresan, co-founder of Wise Step Recruiting. Calin spent about 20 years in hands-on software engineering, cloud architecture, and technical leadership before moving into technical recruitment.

Bosch Global Business Services plans to cut about 510 of its roughly 1,800 jobs in Timișoara by the end of 2030, around 28% of the site. The company announced it on December 10, 2025, and gave two reasons: expected “stagnant demand for services” and the productivity potential that exists in services (Economica.net, December 10, 2025). The roles affected are mainly accounting, administration, HR and support.

Some English-language coverage presented it as an AI story. Bosch’s statement, as published by Economica.net, does not use the word AI. We think the story is still about AI, just not the one the headlines told. We are former software engineers who recruit for engineering and data teams, many of them in western Romania. The question clients ask us is whether this is a warning for engineering headcount. It isn’t, and treating it as one leads to bad hiring and retention decisions.

Key takeaways

  • Bosch Global Business Services plans to cut about 510 of 1,800 jobs in Timișoara and about 3,400 worldwide by the end of 2030. Bosch says the GBS cuts count toward the roughly 13,000 job cuts it announced in September 2025.
  • Bosch’s statement cites “stagnant demand for services” and productivity potential in services. It does not mention AI. The “[AI]” in some English quotes is a journalist’s addition.
  • The cuts fall on standardized work: accounting, administration, HR and support. That matches the ILO’s 2025 finding that clerical jobs are the most exposed to generative AI.
  • Stanford’s August 2026 revision, using U.S. payroll data, shows where AI’s effect lands. Employment of workers aged 22 to 25 in AI-exposed jobs is 19% below where it would be if it had kept pace with less-exposed peers, mostly through less hiring. Experienced workers show no comparable gap.
  • For hiring leaders, the lesson is to audit tasks, not job titles. Automate standardized work, keep hiring specialists, and protect your junior pipeline on purpose.

In this article: What Bosch announced · What Bosch said about AI · Which work is compressed · Why it is not an engineering signal · Timișoara’s talent market · Playbook · Limits

What Bosch announced

Bosch Global Business Services runs internal operations for the group, including accounting, purchasing and HR. Its statement, reported by Economica.net on December 10, 2025, says the division plans to cut about 3,400 jobs worldwide by the end of 2030, about 510 of them in Romania. The Eurofound European Restructuring Monitor logs the same announcement date and the same Timișoara headcount (Eurofound factsheet 203942).

Item What was reported Source
Timișoara cut About 510 positions Bosch statement via Economica.net; Eurofound
Timișoara headcount About 1,800 people Bosch statement via Economica.net
Share of site About 28% Our calculation (510 ÷ 1,800)
Worldwide GBS cut About 3,400 jobs Bosch statement via Economica.net
Deadline End of 2030, done gradually Bosch statement; Eurofound
Functions affected Mainly accounting, administration, HR, support Eurofound
Method Largely process automation, not moving work abroad Eurofound
Support offered Compensation under the collective agreement; moves within the group Eurofound
Wider programme Counts toward the roughly 13,000 cuts announced in September 2025 Bosch statement via Economica.net

The wider programme matters for context. In September 2025 Bosch said it would cut about 13,000 more jobs in its auto-parts (Mobility) business by 2030, mainly in Germany, to close a cost gap of about €2.5 billion (Euronews, September 26, 2025). As of November 9, 2026, we found no public update from Bosch on the Timișoara plan beyond the December 2025 announcement. Timișoara is a small part of a large cost programme driven by the car industry, not a standalone AI experiment.

Bosch Global Business Services Timișoara: positions to be cut by 2030 versus positions kept Stacked bar chart. Of about 1,800 positions at the Timișoara site, about 510, or 28%, are planned to be cut by the end of 2030. About 1,290, or 72%, remain. About 28% of the Timișoara site, by 2030 Bosch Global Business Services, about 1,800 positions 510 About 1,290 kept To be cut by end of 2030 (28%) Remaining positions (72%) Announced Dec 2025; phased in to end-2030
Source: Bosch statement to Economica.net, December 10, 2025; Eurofound factsheet 203942. Percentages are our calculation.

What Bosch said, and what it didn’t

Bosch’s statement gives two reasons, in this order: expected stagnant demand for services, and “the productivity potential that exists in the service environment” (our translation of the Romanian text published by Economica.net). The word AI does not appear in the statement as published. Romania Insider’s English version quotes the phrase as “productivity potentials [AI],” and the square brackets mark a journalist’s addition (Romania Insider, December 11, 2025).

This is not pedantry. A company that says “AI replaced these people” is making a claim about technology. A company that says “demand is flat and we can do the same work with fewer people” is making a claim about volume and process. Bosch made the second one.

Still, the gloss is reasonable. When Eurofound summarises the plan, it says the cuts will happen “largely through process automation.” In shared services in 2026, process automation means invoice matching, ticket triage, document extraction and first-draft reporting. Generative AI and older automation tools now do that work together. So the honest version is this: Bosch is betting that flat demand plus automation, much of it AI-assisted, lets it run its back office with about 28% fewer people in Timișoara. That claim is narrower than “AI is cutting jobs,” and more useful.

Which work AI is actually compressing

AI exposure is highest where the work is standardized, rule-based and document-heavy, which is the profile of the functions Bosch named. The International Labour Organization’s refined 2025 index finds that “clerical occupations continue to have the highest exposure levels.” It also finds that one in four workers globally is in a job with some generative AI exposure, and only 3.3% are in the highest exposure group (ILO Working Paper 140, May 20, 2025).

The ILO adds two points that matter here. First, it expects most jobs to be transformed rather than eliminated, because most occupations mix automatable tasks with tasks that need human input. Second, exposure has risen for some highly digital professional and technical roles. Engineers are not immune. Their exposure sits in specific tasks, not in the whole role.

The table below is our classification, informed by the ILO index and the Stanford data. It sorts work by what it is made of, not by department:

Work type Examples AI exposure What happens in practice
Standardized, high volume Invoice processing, data entry, payroll checks, first-line HR queries High Fewer people per unit of work. Volume moves to automated flows plus human review.
Standardized, judgement at the edges Month-end close, procurement exceptions, reporting Medium Roles shrink and shift toward handling exceptions and controls.
Specialist, ambiguous System design, safety-critical embedded software, incident response, model evaluation Task-level AI speeds up parts of the job. Demand for experienced people holds.
Entry level in any function Junior analysts, junior developers Hiring effect first Less hiring at the entry level. This is where the effect shows up first.

The last row is the uncomfortable one, and it applies to engineering too.

Why this is not a signal about engineering headcount

The best current evidence says AI’s job effect shows up through less hiring of young workers, not through cutting experienced specialists. Stanford’s Digital Economy Lab uses payroll data covering millions of U.S. workers. In its August 2026 revision, it finds that employment of workers aged 22 to 25 in AI-exposed jobs is 19% below where it would be if it had kept pace with less-exposed peers (Brynjolfsson, Chandar and Chen, Stanford Digital Economy Lab, revised August 12, 2026). Experienced workers show no comparable gap. The declines cluster where AI substitutes for human tasks. Where AI complements workers, employment is flat or rising.

Employment gap in AI-exposed occupations by worker group, United States, through June 2026 Lollipop chart. Workers aged 22 to 25 in AI-exposed occupations: employment 19% below trend relative to less-exposed peers. Experienced workers in the same occupations: no comparable gap, shown as zero. AI's job effect lands on hiring, not on seniors Employment gap vs. less-exposed peers, AI-exposed jobs Aged 22–25 −19% Experienced workers No comparable gap −30% −20% −10% 0% U.S. payroll data (ADP) through June 2026
Source: Brynjolfsson, Chandar and Chen, “Canaries in the Coal Mine?”, Stanford Digital Economy Lab, revised August 12, 2026. The authors describe these as early, descriptive indicators.

Bosch’s plan is gradual too: five years, mostly through automation. Bosch has not said how many of the 510 will leave through attrition. Eurofound records compensation packages under the collective agreement and moves within the group, so some people will lose their jobs. For the local market, the likely effect is fewer back-office openings over several years, not one large layoff.

Here is the mistake we want clients to avoid. A leadership team reads “AI cut 28% of Bosch’s Timișoara site” and concludes that its own engineering plan should shrink by a similar amount. That mixes up two kinds of work. Shared-services work has fixed inputs, known rules and measurable output per person, so automation shows up directly as fewer people. Specialist engineering work has unclear requirements, safety and reliability limits, and accountability that cannot be handed to a tool.

In our searches, demand for experienced engineers who can design, review and own AI-assisted systems has not dropped, and those people still take longer to find than most roles we fill. A plan that freezes senior engineering hiring because of a back-office headline leaves you short of exactly the people who make automation work. For how AI changes engineering teams, rather than replacing them, see our guide to AI-native engineering teams.

The risk that does apply to engineering is the junior pipeline. If AI tools absorb the tasks you used to give new graduates, you will hire fewer juniors this year. Three years from now, you will have fewer mid-level engineers to promote.

What it means for Timișoara’s talent market

Timișoara keeps the site. Bosch says the location stays focused on its competitiveness as a strategic centre for the group (our translation of the Bosch statement via Economica.net). The direct effect on hiring here is modest and spread over five years:

  • This is not an engineering talent wave. The affected roles are finance, procurement, HR and support. Some people will retrain into analytics, process automation or data roles, and those candidates are worth a look for operations-adjacent technical roles. But do not plan engineering hiring around this.
  • Romania’s business services sector is large enough to absorb gradual cuts. Deloitte’s 2026 study for ABSL (covering 2025) counts more than 280,000 professionals in the sector, generating €35.5 billion in gross value added, about 10% of Romania’s total gross value added (Deloitte for ABSL Romania, August 11, 2026). About 510 roles over five years is a small share of that.
  • Centres are moving up the value chain. Adrian Baron, president of ABSL Romania, says the sector’s people increasingly work on “technology, data analytics, automation and artificial intelligence” rather than only support services (same Deloitte release). In Timișoara, we expect that to mean more demand for people who build the automation, and less for people who do the work it replaces.

For current pay benchmarks across these roles, see IT salaries in Romania in 2026. For the regional picture, see IT recruitment trends in Eastern Europe in 2026.

A playbook for hiring leaders

If you run a shared-services centre, an engineering team, or both, these five steps turn the Bosch news into decisions:

  1. Audit tasks, not titles. List the recurring tasks in each role and mark which ones are standardized and high volume. A role that is 80% standardized will shrink. A role where they are 20% will mostly speed up.
  2. Phase reductions over years, as Bosch is doing. Our advice: use attrition and internal moves first. A multi-year, gradual plan lets you redeploy people instead of laying them off, and keeps trust with the people you need to keep. Announce the direction early. Silence does more damage to retention than the plan itself.
  3. Hire the people who build and run the automation. Every automated flow needs someone to design it, monitor it and handle exceptions. Those roles, from process engineers to applied AI engineers, are where hiring grows.
  4. Keep hiring specialists on their own plan. Base engineering headcount on product roadmaps and reliability needs, not on headlines about the back office.
  5. Protect a junior pipeline on purpose. Keep a fixed number of graduate and junior seats, and pair them with experienced engineers who review AI-assisted work. It is usually cheaper than competing for mid-level engineers later.

If you are sizing a technical team in western Romania and want a second opinion on which roles to hire, our team build-out service starts with exactly this kind of task-level review.

Where this reading is weakest

  • We rely on Bosch’s statement as reported. The figures come from Bosch’s statement to Economica.net and from Eurofound’s summary. We have not seen Bosch’s internal plan, and it has not published which roles go first.
  • “Mostly automation” is Eurofound’s phrase. Bosch’s own wording is about productivity potential and flat demand. How much of the cut comes from AI, how much from older automation and how much from lower demand is not public.
  • The Stanford data is from the U.S. It covers U.S. payroll data, and the authors call the findings early, descriptive indicators. We use it as the best current evidence on where AI effects land, not as a forecast for Romania.
  • Our view on engineering demand comes from our own searches. It reflects the roles we recruit for, not a market-wide survey. If AI tools absorb more specialist tasks than we expect, that view will need revisiting.

The Bosch cut is real and large for one site. Read correctly, it says that standardized services work is getting cheaper to run. It does not say that engineers are next. Make your hiring decisions on that basis.


Last updated: November 10, 2026. Sources: Economica.net, December 10, 2025; Eurofound factsheet 203942; Romania Insider, December 11, 2025; Euronews, September 26, 2025; ILO Working Paper 140, May 20, 2025; Stanford Digital Economy Lab, revised August 12, 2026; Deloitte for ABSL Romania, August 11, 2026. Retrieved November 9, 2026.