ARPA-H HEARING: Inside the Brain-Connected Hearing Restoration Program That Just Set Solution-Summary Deadlines for June 29 — and Why This Is the First Federal Solicitation to Treat Auditory Cortex as the Implant Target

ARPA-H released solicitation ARPA-H-SOL-26-154, the HEARING program — Hearing Enhancement through ARtificially Intelligent NeurotechnoloGy — earlier this spring and held its Proposers' Day on June 8, 2026. The program will fund development of minimally invasive, brain-connected hearing restoration platforms that bypass the cochlea entirely and write auditory information directly into the auditory cortex. Solution Summaries are due Monday, June 29 at 2:00 PM ET. Full Proposals are due Thursday, August 14 at 2:00 PM ET. The Q&A window closed June 15. The agency will make multiple Other Transaction (OT) awards under a single-prime, multidisciplinary-team structure spanning 4.5 years and three integrated technical areas.

For decades, hearing restoration has meant the cochlear implant — a device first approved by the FDA in 1984 that stimulates the auditory nerve through electrodes threaded into the cochlea. Cochlear implants restore some hearing to roughly one million recipients worldwide. They do not work for the substantial population whose hearing loss involves auditory nerve damage, cochlear nerve aplasia, or central auditory pathway dysfunction. They also have well-documented limitations on speech-in-noise performance, music perception, and bilateral spatial hearing. The HEARING program is the federal government's first concerted bet on what comes after the cochlear implant: direct cortical interfaces that read and write auditory information at the level of the brain.

For context on ARPA-H's broader program portfolio in 2026, see Granted News and our prior coverage of ARPA-H's IGoR program, STOMP, and the agency's $139 million EVIDENT initiative.

The Three Technical Areas, and Why They Have to Move Together

HEARING is structured around three integrated technical areas, and the solicitation is explicit that proposals must address all three. This is unusual even for ARPA-H, which has historically tolerated split TA awards. The agency's stated rationale is that auditory cortex hearing restoration is not a problem that can be solved by any one technology — the brain interface, the sensor, and the decoding algorithm have to be co-engineered from the start.

TA1 — Intracortical Devices. The brain interface itself: recording electrodes, stimulation electrodes, wireless data and power transfer, surgical delivery systems, and the cortical targeting accuracy required to write to auditory cortex without disrupting adjacent functions. The solicitation rules out devices requiring craniotomy or craniectomy as the final form factor. This is not a research surgery. The device has to be implantable through a minimally invasive procedure that hospital-credentialed surgeons can perform in a standard ambulatory setting — a constraint that immediately rules out the Utah array and the modified Neuralink-style implants. What it allows for are emerging stentrode-class technologies, polymer microelectrode arrays delivered through stereotactic burr-hole procedures, and the newer generation of injectable mesh electronics.

TA2 — Dynamic Sound Modulator. The wearable component: a non-implanted device that captures environmental audio, processes it for the auditory features that the cortical interface can write, and transmits to the implant via wireless power and bidirectional data link. This is not a hearing aid in the traditional sense. The DSM is the bottleneck through which all sound passes before it becomes neural signal. Wireless power transfer at sufficient throughput to maintain a closed-loop cortical stimulation device is itself a hard engineering problem, and TA2 is funded to solve it.

TA3 — Auditory Read & Write Algorithms. The AI/ML layer: decoding algorithms that interpret cortical recordings to estimate the user's auditory perception state, and encoding algorithms that translate the DSM's output into stimulation patterns the implanted device can deliver. This is where the program gets fundamentally new. Existing cochlear implant signal processing strategies — continuous interleaved sampling, spectral peak strategies, fine-structure processing — were designed for peripheral electrical stimulation of a quasi-tonotopic structure. Cortical auditory processing is not tonotopic in the same way. The TA3 algorithms have to be developed from auditory neuroscience first principles, validated against electrophysiology, and personalized to the individual recipient's cortical architecture.

The integration requirement matters because each TA has design constraints that flow from the others. The DSM's processing budget is set by the implant's stimulation channel count and bandwidth. The algorithms' latency budget is set by the wireless link's throughput. The implant's power budget is set by what the DSM can wirelessly deliver. A team that nails one TA without integrating with the others will produce a research result, not a clinical platform.

The Phase Structure: 18 / 24 / 12

The 4.5-year effort is divided into three phases.

Phase 1 — R&D, 18 months. Develop integrated prototypes, characterize cortical recording and stimulation in animal models, validate decoding and encoding algorithms in non-human primates or other appropriate models, and demonstrate the wireless DSM-implant communication architecture under benchtop conditions. End of Phase 1 must produce an integrated system ready for pre-clinical safety and efficacy validation.

Phase 2 — Pre-Clinical, 24 months. Conduct GLP-compliant safety and efficacy studies in animal models. Engage FDA on regulatory pathway — almost certainly through the Breakthrough Devices Program and a Pre-Submission process. Develop the manufacturing protocols, sterilization processes, and supply-chain qualifications required for first-in-human studies. End of Phase 2 must produce an Investigational Device Exemption-ready package.

Phase 3 — Clinical / First-in-Human, 12 months. Conduct first-in-human studies under IDE in carefully selected patient populations. The solicitation does not specify enrollment targets at this phase, which is appropriate for an early feasibility study. The expectation is that successful Phase 3 work positions the platform for a follow-on commercialization vehicle — either a private fundraise, a strategic acquisition, or an extension of the OT into expanded clinical studies.

The phase structure is aggressive. 4.5 years from kickoff to first-in-human is faster than the historical neuro-prosthesis development timeline by roughly a factor of two. ARPA-H is funding this on the assumption that the underlying technology platforms — flexible electrode arrays, intracortical wireless devices, AI-based neural decoding — have matured enough in the past five years that integration rather than invention is the binding constraint.

Eligibility and the Single-Prime Requirement

The solicitation requires a single prime awardee with a multidisciplinary team covering all three technical areas. Universities, non-profits, small businesses, and other organizations are eligible to serve as prime. Federally Funded Research and Development Centers cannot serve as prime performers without explicit prior approval. Government entities and certain foreign organizations are ineligible.

The single-prime structure is a design choice ARPA-H has used before, most notably in IGoR and STOMP. It pushes program management complexity onto the prime, but it ensures that there is one entity accountable for delivery and one set of contracting terms governing IP, data rights, and milestone payments. Teams should select their prime based on which entity can credibly manage a 4.5-year, three-TA, multi-institutional program — not on which entity has the most prominent scientific contribution.

Submission limits are tight: one submission as prime, two as sub-proposer. This means a research group cannot hedge by serving as prime on one team and as a co-PI on another competing prime. Teams should formalize their structure early.

What This Means Against the Cochlear Implant Industry

The cochlear implant market is dominated by three vendors: Cochlear Ltd (Australia), Sonova (Switzerland, parent of Advanced Bionics), and MED-EL (Austria). The combined market is roughly $2.2 billion annually and has grown at single-digit rates for the past decade. None of the three has a public roadmap for direct cortical hearing restoration. The HEARING program is, in effect, the U.S. government's bet that the next generation of hearing restoration will come from the brain-machine interface industry rather than from the incumbent cochlear implant companies.

This has implications for the applicant pool. The teams most credibly positioned to compete are not the existing cochlear implant vendors — they have neither the cortical neuroscience expertise nor the BMI hardware portfolios. They are the teams that have spent the past decade building intracortical interfaces for motor restoration (Synchron, Precision Neuroscience, BrainGate Consortium, Blackrock Neurotech) and the academic auditory neuroscience labs (UCSF, UCSD, MIT, Pittsburgh, Hopkins, Iowa, NYU) that have characterized auditory cortical processing at the resolution required. The most likely winning team structure is a small-business or non-profit prime that consolidates one of those BMI hardware platforms with two or three of those auditory neuroscience labs and an AI/ML lab with neural decoding experience.

Regulatory and Commercialization Considerations

The HEARING platform will be a Class III medical device requiring Premarket Approval. The FDA pathway is non-trivial. The Breakthrough Devices Program designation is essentially required to make the timelines work — Breakthrough status accelerates review, allows for sprint-style FDA engagement, and unlocks Medicare's NCD pathway for transitional coverage. Teams should plan for a Q-Sub meeting in Phase 1, a Pre-Submission in early Phase 2, and an IDE submission in late Phase 2.

CMS coverage is the longer-term question. Cochlear implants are covered by Medicare under a longstanding NCD. Direct cortical hearing restoration is novel enough that a new Coverage with Evidence Development pathway is likely. Teams that build the clinical evidence requirements into Phase 3 study design will be substantially better positioned for the commercialization phase than teams that treat regulatory and reimbursement as a downstream concern.

How to Read the Solicitation This Week

Three things matter most for teams preparing Solution Summaries by June 29.

First, the TA1 form-factor constraint. The prohibition on craniotomy / craniectomy final device forms is not a soft preference. ARPA-H will reject Solution Summaries that propose research-only implants and treat clinical translation as a downstream problem. The Solution Summary needs to make clear how the team will deliver a minimally invasive implantable device by end of Phase 2.

Second, the integrated end-to-end demonstration. Solution Summaries that propose strong work on one or two TAs without a credible integration plan will not make it past initial review. The agency has been explicit that integration is the program — the TAs are not parallel tracks.

Third, the first-in-human commitment. ARPA-H is not funding a 4.5-year basic research program. The Phase 3 first-in-human studies are the program's goal, and the Solution Summary needs to demonstrate that the team has the regulatory, clinical, and ethical infrastructure to deliver a first-in-human study by Q4 of year 4.5. Teams without an academic medical center clinical site partner with neurosurgery, audiology, and IRB infrastructure are not viable.

Adjacent Federal Programs and Funding Stack Options

HEARING does not exist in isolation. The Department of Veterans Affairs operates one of the largest cohorts of patients with combat-related hearing loss in the world and has its own neurotechnology research portfolio through ORD and the Rehabilitation R&D Service. NIH's National Institute on Deafness and Other Communication Disorders has ongoing R01 and R21 cycles in auditory cortical processing. The DoD's Hearing Restoration Research Program through CDMRP funds related basic science. Successful HEARING teams should be coordinating with all three, both for complementary funding and for clinical site access.

For teams that don't win HEARING in this cycle but build credible Solution Summary packages, those packages are reusable. NSF's emerging neurotechnology investments, the National Institute of Biomedical Imaging and Bioengineering, and the BRAIN Initiative's R01 and U-mechanism programs all fund adjacent work. ARPA-H itself is likely to issue follow-on solicitations as the program portfolio matures.

Practical Steps for Granted Users

If you are a neurotechnology team, search Granted for the full HEARING package plus the BRAIN Initiative R21, R01, and U-mechanism cycles that complement it. If you are a clinical site building neurosurgery infrastructure, browse our medical device grant database for the trial-infrastructure supplements that pair with this program. If you are an auditory neuroscience lab looking for a prime partner, search past ARPA-H awardees for organizations that have managed prior OT awards at this scale.

The Solution Summary deadline is twelve days away. The Full Proposal deadline is fifty-eight days away. The first-in-human study, if your team wins, is four years away. ARPA-H has just set the clock.