Bloomberg Data Science Ph.D. Fellowship 2027–2028 (Historical Reference)

Bloomberg Data Science Ph.D. Fellowship 2027–2028: historical reference

The 2027–2028 application cycle is closed. The Bloomberg Data Science Ph.D. Fellowship committee announced a final submission deadline of September 25, 2026, at 11:59 PM Anywhere on Earth (AoE). The call is no longer open, and the official material reviewed for this page does not announce a later cycle. This entry records the closed call and its verifiable requirements; it is not an invitation to apply now.

The official Bloomberg fellowship page is Bloomberg’s Data Science Ph.D. Fellowship page. In the page version available for review, Bloomberg still describes the prior 2026–2027 call and its earlier deadline. The 2027–2028 announcement was published by the Bloomberg Data Science Ph.D. Fellowship committee through the ACL Member Portal and points applicants back to Bloomberg’s official fellowship page. Bloomberg’s own Overleaf proposal template also names the 2027–2028 call and links the matching Bloomberg2027 submission portal. These primary materials confirm a new cycle existed, even though the main program page currently presents older cycle details.

What is confirmed for the closed call

The committee announcement describes the fellowship as support for outstanding doctoral researchers advancing data science, artificial intelligence, and their applications. It presents both financial support and opportunities to engage with Bloomberg. The announcement does not state an award amount or itemize tuition, stipend, internship duration, renewal conditions, or other benefit values. Bloomberg’s accessible program page lists terms for 2026–2027, but those figures and conditions cannot safely be assigned to the later call without a current-cycle statement. For that reason, this archive does not describe the 2027–2028 award as a particular dollar amount or guarantee a particular tuition package.

This distinction matters when comparing funding offers. A previous cohort’s public description can be useful background, but it is not proof that the same package applied to this call. Applicants who submitted to the closed round would have needed to rely on the actual call documents and communications they received for award details. Readers assessing a future announcement should check its current terms directly and compare the stated tuition support, stipend, payment schedule, duration, renewal rules, and internship obligations before making financial plans.

Research areas named in the announcement

The committee’s 2027–2028 notice lists the following areas of interest. It says the call covers these topics, not that the list guarantees selection or excludes adjacent work:

• Agentic AI, tool use, and agent infrastructure.
• Agent evaluation and LLM judges.
• Code generation, semantic parsing, and text-to-API.
• Efficient models, distillation, and inference.
• Human-AI collaboration, data annotation, and data-centric AI.
• Information retrieval, retrieval-augmented generation (RAG), and agentic search.
• Information synthesis, summarization, and generation.
• Knowledge graphs, structured reasoning, and neuro-symbolic reasoning.
• LLM post-training, domain adaptation, reasoning, and alignment.
• Multimodal models and document intelligence.
• Time-series modeling, forecasting, and AI for finance.
• Trustworthy, responsible, and secure AI.

A useful way to read this list is as a description of research fit rather than a checklist of required keywords. A proposal in information retrieval, for example, would still need to identify a concrete technical problem and explain how its approach could be evaluated. The same applies to systems for agents, document understanding, forecasting, or safe model behavior. The committee notice establishes the subject areas; it does not state a preferred method, benchmark, or expected result for any of them.

Eligibility: what the record establishes

The 2027–2028 committee notice characterizes the intended applicants as Ph.D. students and refers to outstanding doctoral research in relevant areas. It does not give a complete eligibility schedule in the announcement itself. In particular, the materials verified here do not confirm a 2027–2028 rule for the applicant’s full-time enrollment status, expected graduation date, location or work authorization, other fellowships, outside employment, kickoff attendance, or internship participation.

Bloomberg’s program page does describe detailed criteria for the prior 2026–2027 academic-year call. Those rules are cohort-specific. They should not be copied into a 2027–2028 eligibility checklist as if Bloomberg had reaffirmed them. The same caution applies to an aggregator, old application link, search snippet, or a previous cohort’s terms: it may be useful for locating relevant questions, but it is not sufficient evidence that the conditions applied to this round.

This archive therefore does not claim that every doctoral student could apply, nor does it invent a graduation cutoff or authorization requirement. Anyone reviewing the closed round should use the official instructions and direct correspondence that accompanied their application. Anyone preparing for a later round should wait for its own eligibility language and confirm any situation-specific question with Bloomberg’s program organizers. This is especially relevant where an award combines university enrollment with a corporate internship, because institutional status, employment permission, and work authorization can depend on an individual’s circumstances.

Application materials and submission route

Bloomberg’s own Overleaf template is explicitly titled for the 2027–2028 Data Science Ph.D. Fellowship application. It says applicants were to prepare a research proposal using the template’s formatting requirements. The proposal could contain no more than two pages of content, with references allowed on additional pages. Bloomberg warns in the template that proposals that do not follow its specifications would not be reviewed. It also says the application should be a PDF and that appendices or supplementary material such as data or code were not accepted.

The template directs applicants to submit the proposal through the Bloomberg2027 SoftConf application portal, together with a CV and referee names. The committee’s call notice adds that recommendation letters were also due by the deadline and that applicants were responsible for making sure referees submitted them. Taken together, those primary materials establish these steps for the closed call:

1. Review the fellowship call and determine whether the research proposal fits its stated areas.
2. Prepare a CV and a technical proposal no longer than two pages of content, excluding references.
3. Use Bloomberg’s published Overleaf template and preserve its required layout and format.
4. Submit the proposal PDF through the Bloomberg2027 SoftConf portal and provide the referee information requested there.
5. Ask the advisor or other referee to submit the requested recommendation letter, then check that the materials arrived by the cutoff.

Those are archived submission instructions. Since the deadline passed, the portal and template should not be treated as evidence that Bloomberg is still taking applications. A later round could use a different portal, document set, formatting file, or deadline; nothing in the closed call confirms that these steps will repeat.

Deadline and archive status

The short deadline value for this page is 2026-09-25, matching the date in the fellowship committee’s 2027–2028 announcement. The full notice gives the cutoff as 11:59 PM AoE and applies it to applications and reference letters. This is a fixed deadline, not a rolling window. No official statement reviewed here says that late submissions remained eligible or that applications would continue until places were filled.

The call is marked as historical because the published deadline has passed and no later round is announced in the official materials reviewed. The historical flag is not a claim that Bloomberg has ended the fellowship permanently. It means only that this particular application round has closed. It preserves the genuine deadline so that future readers can distinguish the archived call from an open opportunity.

The official program page may continue to show an “Apply Now” link or information from an older cohort. A visible link alone does not demonstrate that a form accepts submissions for a current round. Check the date and academic year named in the call itself. The main Bloomberg page reviewed for this update still refers to the 2026–2027 academic year and its earlier deadline; the fellowship committee announcement and Bloomberg’s 2027–2028 template are the sources that identify the later cohort. If these official materials are updated again, the newest cohort-specific instructions should take precedence over this archive.

Questions about this historical call

Can I apply to the 2027–2028 fellowship now? No. The published deadline was September 25, 2026, at 11:59 PM AoE. The committee notice gives no late-submission or continuous application route.

How much funding did the 2027–2028 call offer? The verified committee announcement confirms financial support but does not state a figure or itemize benefits. This page leaves the amount unspecified rather than reusing prior-cycle figures as current facts.

Who was eligible? The announcement describes the opportunity for Ph.D. students working in relevant areas. It does not supply the detailed enrollment, graduation, work authorization, employment, or internship conditions. The older rules on Bloomberg’s program page belong to the prior academic-year call and are not presented here as confirmed for this one.

Which topics were included? The committee notice names agentic AI and infrastructure, evaluation, code generation, efficient models, human-AI collaboration, retrieval and RAG, synthesis, knowledge representation, post-training, multimodal and document intelligence, time-series and finance, and trustworthy AI. The full list appears above.

What did applicants submit? Bloomberg’s 2027–2028 template specifies a CV, referee names, and a research proposal PDF of at most two content pages, excluding references. Applicants had to follow the template. The committee notice says reference letters were due by the same deadline and that applicants were responsible for ensuring their arrival.

When will the next round open? No later cycle is confirmed in the official materials reviewed. Do not infer a recurring date from this deadline. Consult Bloomberg’s official fellowship page for a dated announcement and the terms of any future call.

Official sources

• Bloomberg Data Science Ph.D. Fellowship page — official Bloomberg program page; the version available during this update still described the earlier call.
• Bloomberg committee announcement through the ACL Member Portal — identifies the 2027–2028 cohort, its topic list, and the September 25, 2026 deadline.
• Bloomberg-authored application template on Overleaf — names the 2027–2028 application, specifies proposal formatting, and links the Bloomberg2027 portal.

Use Bloomberg’s official fellowship page for any later announcement and verify that the page names the intended academic cycle. The material above is an archive of the closed 2027–2028 call, with unspecified award and eligibility details deliberately left unasserted.