Which banks are hiring machine learning and data science roles? 88 open roles at 33 banks.

Which banks have the most ML & data science openings?

Data as of Sep 24, 2026 – Oct 14, 2026 · 33 banks · 88 postings · each bank's own careers site. A job posting is evidence that a bank is recruiting, not of what it has deployed; for that, read the bank's page. Counts are AI-titled postings listed on the profile, not headcount, and a profile lists at most 12 postings per bank.

What ML & data science titles are banks posting?

Data as of Sep 24, 2026 – Oct 14, 2026. Each title links to the posting on the bank's careers site; postings close without notice.

Where are the ML & data science jobs?

Data as of Sep 24, 2026 – Oct 14, 2026. First city named on each posting; 6 postings in this family name no city. All cities →

How does ML & data science compare with the other role families?

How are these counts built, and what do they not show?

Each bank page on this site carries an “Open AI roles” block: the postings whose title names AI, machine learning, data science, generative AI, LLMs, agents or MLOps on the bank's own careers site on the date shown, found by searching that site and keeping only titles that are about AI. 58 of the 140 bank pages carry such a block as of Sep 24, 2026 – Oct 14, 2026; the other 82 have not been surveyed yet, or their careers sites could not be searched, so a bank missing here is not a bank that is not hiring. A profile lists at most 12 postings, so a count of 12 means “at least 12” and is shown as 12+.

Role families are assigned from the title alone with fixed word lists (responsible AI & governance: titles naming responsible AI, AI governance, AI risk, model risk, AI compliance, supervision, controls, validation or ethics; AI security: titles naming security, cyber, identity, threat, red team, offensive security, DLP or data protection alongside AI; AI research: titles naming AI research, research scientists or fellows, and applied machine-learning scientists; agentic AI & LLM: titles naming agents, agentic AI, LLMs, generative AI, RAG, prompt or context engineering, conversational AI, copilots or chatbots; ML & data science: titles naming data science, machine learning, ML, MLOps, modelling, quantitative analytics, data engineering or data architecture alongside AI; AI product & strategy: titles naming product, programme, strategy, transformation, enablement, adoption, go-to-market, consulting, analysis or AI leadership (head of AI), plus AI internships; AI engineering: titles naming engineers, architects, developers, SREs, DevOps, platforms, infrastructure, cloud, integration or automation alongside AI); a title can sit in more than one family, so family counts add up to more than the number of postings. Seniority is also read from the title: VP, SVP, director, executive director, managing director and “head of” count as leadership (at US banks a vice president is often a mid-level grade); principal, lead, staff, distinguished and manager titles are the middle band; everything else, interns included, is the third. Locations are the first city a posting names; postings that name only a country, “2 locations” or nothing are left out of the city tables.

Postings are a snapshot: they close, reopen and get re-titled, and the same requisition can appear in several cities. Counts are not headcount, not budgets and not a measure of how far a bank's AI programme has gone; a bank that hires through agencies, internal transfers or untitled “software engineer” requisitions will look smaller here than it is. Where a posting carries no date, the careers site did not show one. Snapshots older than 45 days are flagged as stale; each bank's page shows the same list with its own date, and the AI index is the place for what banks have actually put on the record.