Google Academic Research Awards 2026: Up to $100,000 as an Unrestricted Gift to Your University for Early Fault-Tolerant Quantum Algorithms or Quantum Computing Security — Proposals Close 7 August 2026

Google Academic Research Awards 2026: Up to $100,000 as an Unrestricted Gift to Your University for Early Fault-Tolerant Quantum Algorithms or Quantum Computing Security — Proposals Close 7 August 2026

Google Research runs open calls for proposals under the Google Academic Research Awards (GARA) programme, and the 2026 cycle is unusually focused. Instead of spreading across a dozen computing themes, Google has published exactly two global requests for proposals this round, and both sit inside quantum computing: Early Fault-Tolerant Quantum Algorithms & Applications, and Quantum Computing Security. Each award is worth up to $100,000, delivered as an unrestricted gift to the applicant’s university or degree-granting research institution. Applications opened on 26 June 2026 and close on 7 August 2026 at 11:59:59pm AoE (UTC-12), with decisions announced by 30 October 2026.

That narrow scope is the most important thing to understand before you spend a week writing. If your work does not touch early fault-tolerant quantum algorithms or the security of quantum computing systems, this particular round is not for you — and waiting for a future cycle with a different theme is the rational move. If your work does sit in one of those two areas, this is a fast, low-bureaucracy route to six figures of genuinely unrestricted money, with no indirect-cost skim and no multi-year reporting obligation.

Key Details at a Glance

What the Award Actually Gives You

The headline is up to $100,000. Google states it will consider larger amounts for exceptional proposals that make a clear and compelling case for the increased funding, so the ceiling is soft rather than absolute — but you should treat $100,000 as the planning number unless you have a specific, defensible reason to ask for more and can articulate it in the proposal.

The structure matters as much as the number. These awards are made as unrestricted gifts to the institution, not as contracts or cost-reimbursement grants. In practice that means three things:

1. No indirect costs. Google is explicit that funding awarded in any form should not be used for overhead or indirect expenses. Your research office will need to accept the gift on those terms, which most institutions do for corporate gift funding, but it is worth confirming internally before you submit rather than after you win.
2. Broad spending latitude. Unrestricted gift funding is typically the most flexible money a lab can hold — student stipends, compute, travel, equipment — subject to your own institution’s gift-administration rules rather than a funder’s line-item budget.
3. Light-touch obligations. For awards structured as unrestricted gifts, Google notes there are no legal requirements once a project is selected. The data-sharing section of your proposal is a statement of intent, not a binding covenant.

There is an important exception to that last point. Some GARA research topics are not awarded as unrestricted gifts — usually those that require the use of a specific Google product, methodology, or other constraint. For those, open-sourcing the software, models, or other intellectual property developed during the project becomes a mandatory condition of the award unless a separate agreement says otherwise. Read the individual RFP page for the call you are targeting and confirm which regime applies to it before you make promises to collaborators about what will and will not be published.

The Two 2026 Calls, and Which One Fits You

Early Fault-Tolerant Quantum Algorithms & Applications

This call is aimed at the near-term regime where a quantum computer has real error correction but only a relatively small number of logical qubits — enough, potentially, to attack problems that are classically intractable, but nowhere near the resource budgets assumed by textbook fault-tolerant algorithms. Google identifies three primary areas of interest:

• Low-resource novel algorithms — new quantum algorithms that are actually feasible on early fault-tolerant hardware, rather than asymptotically attractive but resource-hungry.
• Early fault-tolerant practical applications — mapping concrete real-world problems onto known algorithms in a way that respects early-era constraints.
• Resource reduction and overhead optimisation — compilation, error mitigation, and other techniques that shrink the gap between what an algorithm needs and what the hardware can supply.

Google names life sciences and health, climate and sustainability, and energy and materials sciences as application domains of particular interest. If your proposal can land a credible logical-qubit and gate-count estimate against a specific problem instance in one of those domains, you are speaking the reviewers’ language.

Quantum Computing Security

The second call is narrower and, for most applicants, less crowded. It centres on the classical-to-quantum interface — the layer where high-level instructions get translated into physical control pulses — and on protecting quantum computing efforts from IP theft and practical attack. Areas Google flags include:

• Real-time verification of pulse-level instructions against gate specifications
• RF and microwave emission analysis, and defences against it
• Secure reset protocols to prevent data remanence between jobs
• Multi-tenant isolation on cloud-accessible quantum processors
• Novel attack vectors across different qubit modalities
• Quantum-aware defensive tooling and auditing frameworks
• Security of third-party hardware components, including hardware trojans
• Detecting adversarial interference without disrupting the computation itself

This is a hardware-security research agenda applied to quantum stacks, and it rewards people who already think in terms of side channels, trust boundaries, and supply-chain integrity. If you are a systems-security researcher who has been looking for a way into quantum, this call is a more realistic entry point than the algorithms call.

Eligibility and the One-Proposal Rule

Eligibility is tight and easy to check. You must be a professor — assistant, associate, or higher — at a degree-granting research institution, currently advising students and conducting research in computing or technology. Your proposal must relate to computing or technology, which the two quantum calls obviously satisfy.

Co-PIs must generally meet the same criteria as the lead PI, with one stated exception: postdoctoral researchers may serve as co-PIs. That is a meaningful concession if you want to give a strong postdoc visible leadership on a funded project.

Two limits catch people out:

• One proposal per applicant per round, counted whether you appear as PI or co-PI. Being named on a colleague’s submission uses up your slot. Coordinate across your department before anyone hits submit.
• Maximum two PIs per proposal. Large consortium framings do not fit this programme.

The programme is a global open call and is not restricted to a single country, but institutional eligibility and terms are governed by the programme’s published terms page, which is the authority if your institution has an unusual structure.

Submission Requirements: Format Is Not Optional

Google publishes precise formatting rules, and with a high-volume single-round competition there is no upside to testing how strictly they are enforced.

• One PDF. Combine the proposal and CV or CVs into a single PDF file.
• Filename convention. [First Initial][Last name]-2026 — for example, JDoe-2026.
• Typography. Single-spaced, 1-inch margins, minimum 10-point font.
• Length. Five pages maximum for a single-PI submission; seven pages maximum when a co-PI CV is included.
• Overview section. Three pages maximum, covering the title, contact information, an abstract, your research goals, a description of the proposed work with expected outcomes and results, a discussion of how the research relates to prior work (including your own), an explanation of your qualifications to conduct it, and your data-sharing intentions.
• CVs. Four pages maximum in total — two for the PI, and up to two more for an optional co-PI.
• References and budget are submitted separately through the application tool, not inside the PDF.

Submission runs through Google’s Submittable portal at google.submittable.com/submit. Create your account early; a portal registration problem discovered on 7 August is a lost year.

How Proposals Are Judged

Google lists the evaluation dimensions plainly:

• Faculty merit — your record of research accomplishment and your potential for responsible innovation.
• Research merit — alignment with Google’s stated interests, the innovation in the idea, and its likely impact on the field.
• Proposal quality — clarity, organisation, and whether the plan is feasible to execute.
• Alignment with Google’s AI Principles.
• Adherence to a specified product or methodology, where the RFP requires one.

Google also publishes its own guidance on strengthening a proposal, and it reads like a list of the failure modes reviewers see most: define the problem clearly; describe specific outcomes rather than vague ambitions; differentiate your contribution from existing work; outline your approach including risks and mitigation strategies, even where not all answers are known; contextualise the work against your existing funding and broader research goals; and make the content accessible to non-experts.

That last instruction deserves weight. A reviewer reading a stack of quantum proposals in a compressed window will not reconstruct your argument from dense notation. Lead each section with a plain sentence stating what you will do and why it matters, then give the technical detail.

Timeline and What Happens After You Submit

AoE — Anywhere on Earth — is UTC-12, which means the deadline falls later in absolute terms than most people’s local 7 August midnight. Do not use that as planning slack; portal congestion in the final hours is a real risk.

Between the deadline and late October you should expect silence. Google states plainly that, due to the high volume of applications received, you may not receive feedback on your proposal. Budget for a binary answer with no reviewer comments, and do not build a resubmission strategy that depends on knowing why you lost.

Common Mistakes to Avoid

• Submitting off-theme work. The single most common waste of effort in a themed corporate RFP is a strong proposal in an adjacent area. This round is quantum algorithms in the early fault-tolerant regime, or quantum computing security. Nothing else.
• Burning your one slot as a courtesy co-PI. Check whether a colleague has already listed you before you start writing your own.
• Budgeting overhead. Indirect costs are excluded. A budget that assumes your institution’s standard recovery rate will need rebuilding, and may cause a conversation with your research office at exactly the wrong moment.
• Ignoring the page and filename rules. Five pages, seven with a co-PI CV; references and budget go in the tool, not the PDF; the filename follows the stated pattern.
• Overpromising on open source without checking the award structure. Whether open-sourcing is a mandatory condition depends on how the specific topic is awarded. Confirm this on the individual RFP page.
• Writing only for specialists. Google explicitly asks for accessibility to non-experts. Assume at least one reader is outside your sub-field.

Frequently Asked Questions

Is this open to researchers outside the United States? Yes — Google describes these as global open calls for proposals. Confirm institutional specifics against the programme’s terms page.

Can a postdoc apply? Not as lead PI. The lead must be a professor at assistant rank or above. A postdoc may serve as a co-PI.

Can I ask for more than $100,000? Google says it will consider larger amounts for exceptional proposals that demonstrate a clear and compelling rationale for the increased funding. Treat that as a high bar rather than a routine option.

Can the money cover my institution’s overhead? No. Funding awarded in any form should not be used for overhead or indirect costs.

Will I get reviewer feedback if I am unsuccessful? Probably not. Google states that because of application volume you may not receive feedback.

Does receiving an award create a formal collaboration with Google Quantum AI? The published materials describe a competitive funding programme administered through Google Research; they do not specify a guaranteed collaboration structure. Do not assume hardware access or joint-project status unless it is confirmed to you in writing.

Preparation Strategy and Next Steps

With a 7 August deadline and a five-page limit, this is a proposal you can realistically write well in a focused week — but only if you already have the substance. The strongest submissions in a programme like this are usually not new ideas invented for the call; they are an existing line of work, already partly derisked, framed against Google’s stated areas of interest.

A practical sequence:

1. Read the specific RFP page for your call — the algorithms call and the security call are separate pages with distinct scopes, and they name the sub-areas Google actually wants.
2. Confirm your slot with anyone who might list you as a co-PI.
3. Talk to your research or development office about accepting an unrestricted corporate gift with no indirect-cost recovery. Get that cleared before you submit.
4. Draft the three-page overview first, in plain language, then compress. Cutting a six-page draft to three produces a sharper document than padding a two-page one.
5. Make the outcome concrete. State what will exist at the end that does not exist now — an algorithm with a specific resource estimate, an attack demonstrated on a named interface layer, a tool, a dataset.
6. Register on Submittable early and prepare the references and budget in the tool separately from the PDF.

Official Links and Contact

• Programme overview: https://research.google/programs-and-events/google-academic-research-awards/
• Early Fault-Tolerant Quantum Algorithms & Applications call: https://research.google/programs-and-events/early-fault-tolerant-quantum-algorithms-applications/
• Quantum Computing Security call: https://research.google/programs-and-events/quantum-computing-security/
• Terms: https://research.google/programs-and-events/google-academic-research-awards/terms-for-google-academic-research-awards/
• Past recipients: https://research.google/programs-and-events/google-academic-research-awards/google-academic-research-award-program-recipients/
• Submission portal: https://google.submittable.com/submit
• Questions: [email protected]

Details on this page are drawn from Google Research’s published programme and RFP pages as of 26 July 2026. Where a figure is not confirmed by Google — such as the total number of awards to be made in this round, or the acceptance rate — it is not stated here. Verify deadlines and requirements against the official pages before you submit, since a corporate programme can revise its terms without a formal amendment notice.