Gravitational Waves from Colliding Black Holes Could Enable Detection of Dark Matter, Study Finds

Roughly 27 percent of the universe is made of dark matter — a substance that interacts gravitationally but has never been directly detected in a laboratory [1]. For decades, physicists have built increasingly sensitive underground detectors, pointed gamma-ray telescopes at the sky, and analyzed the cosmic microwave background, all without a confirmed signal. Now, two independent research groups argue that gravitational waves from colliding black holes could offer an entirely different path to detection — one that probes dark matter candidates beyond the reach of conventional experiments.

The proposals are scientifically serious. But they also arrive at a moment when next-generation gravitational wave observatories are competing for billions of euros in public funding, raising questions about whether the dark matter promise is a genuine scientific opportunity or a convenient justification for expensive new instruments.

The Two Studies

The first study, published in May 2026 by a team led by MIT postdoctoral researcher Josu Aurrekoetxea with collaborators Soumen Roy (Université Catholique de Louvain), Rodrigo Vicente (University of Amsterdam), Katy Clough (Queen Mary University of London), and Pedro Ferreira (Oxford University), developed numerical simulations to predict how gravitational waveforms would differ if merging black holes passed through a region of dark matter rather than empty space [2]. They applied their model to 28 of the clearest gravitational-wave signals from LIGO-Virgo-KAGRA's first three observing runs. Twenty-seven signals were consistent with mergers in vacuum. One — GW190728, produced by a binary with roughly 20 solar masses — showed possible signs of a dark matter imprint [2].

The second study, published in Physical Review Letters in November 2025 by Rodrigo Vicente, Theophanes K. Karydas, and Gianfranco Bertone at the University of Amsterdam's GRAPPA center, took a different approach [3]. Rather than analyzing existing data, they built what they describe as "the first fully relativistic framework" for modeling how dark matter environments affect extreme mass-ratio inspirals (EMRIs) — systems where a stellar-mass black hole spirals into a supermassive black hole over months or years, completing hundreds of thousands of orbital cycles [4]. These prolonged signals, detectable by future space-based observatories like LISA (launch planned for 2035), would carry accumulated imprints of any surrounding dark matter.

What Dark Matter Candidates Can This Method Probe?

The gravitational-wave approach targets a fundamentally different class of dark matter candidates than the underground detectors that have dominated the field for two decades.

The MIT-led study focuses on light scalar particles — proposed dark matter candidates with masses many orders of magnitude lighter than electrons — that would behave as coordinated waves near black holes rather than individual particles [2]. This category includes ultralight bosons and axion-like particles with masses in the range of roughly 10⁻²⁰ to 10⁻¹⁰ electronvolts (eV).

This mass range sits far below the sensitivity window of direct-detection experiments like LUX-ZEPLIN (LZ) and XENONnT, which are optimized for weakly interacting massive particles (WIMPs) in the GeV range. The LZ experiment, using a 10-tonne liquid xenon detector at the Sanford Underground Research Facility in South Dakota, reported in December 2025 that it found no evidence of WIMPs between 3 and 9 GeV after 417 live days of data collection — setting world-leading sensitivity limits above 5 GeV [5][6]. XENONnT and PandaX-4T have produced comparable constraints in overlapping mass ranges [5].

In other words, the gravitational-wave method and direct-detection experiments are not in direct competition. They probe different regions of the dark matter parameter space. The ultralight regime that gravitational waves can access is largely unconstrained by laboratory experiments, making the approach complementary rather than redundant.

Signal Requirements and Detection Rates

How many merger detections would be needed to produce a statistically significant dark matter signal?

The MIT team is candid about the current limitations. Aurrekoetxea stated that the GW190728 finding "does not have sufficient statistical significance to claim a detection" [2]. The study serves more as proof of concept — demonstrating that existing gravitational-wave data can be screened for dark matter imprints — than as a detection claim.

The detection rate from current observatories has grown dramatically. LIGO-Virgo-KAGRA's fourth observing run (O4), which concluded in November 2025, detected roughly 250 merger events over about 30 months — a rate approaching several per week [7]. This represents an enormous increase from early operations: O1 (2015–16) detected just 3 events, O2 (2016–17) detected 8, and O3 (2019–20) detected 79 [7].

Source: LIGO Scientific Collaboration

Data as of Nov 18, 2025CSV

But raw detection numbers are only part of the equation. The critical variable is signal-to-noise ratio (SNR). Current LIGO detectors, with 4-kilometer arms, produce SNR values that allow parameter estimation but may not be sufficient to disentangle subtle dark matter effects from other environmental factors. Next-generation detectors — the Einstein Telescope in Europe and Cosmic Explorer in the United States — would improve sensitivity by up to a factor of eight across the relevant frequency band and extend coverage below 10 Hz [8][9]. Cosmic Explorer's planned 40-kilometer arms would increase sensitivity by more than an order of magnitude over LIGO [9].

The Amsterdam group's EMRI-focused approach depends on LISA, the European Space Agency's planned space-based gravitational wave observatory, which would track signals over months to years rather than seconds. The accumulated phase shifts from dark matter interactions over millions of orbital cycles could produce detectable signatures even at moderate SNR [4].

How Does This Compare to Other Dark Matter Search Methods?

Gravitational waves occupy a specific niche in the dark matter search landscape. To understand where they fit:

Direct detection experiments (LZ, XENONnT, SuperCDMS, ADMX) look for dark matter particles scattering off atomic nuclei or coupling to electromagnetic fields. They excel in the WIMP mass range (roughly 1 GeV to 10 TeV) and, for axions, in a narrow band around 1–40 microelectronvolts. They cannot access the ultralight regime below ~10⁻¹⁰ eV [5][10].

CMB measurements from the Planck satellite constrain dark matter annihilation cross-sections at early cosmological times but are insensitive to dark matter's local distribution around astrophysical objects [11].

Gamma-ray telescopes like Fermi-LAT search for annihilation products from dark matter in galactic halos and dwarf galaxies. They set strong limits on WIMP annihilation for masses from a few GeV to tens of TeV but do not probe ultralight candidates [11].

Gravitational lensing observations, including the Bullet Cluster, provide evidence for dark matter's existence and constrain its self-interaction cross-section but cannot determine particle properties [1].

The gravitational-wave method's genuine advantage lies in the ultralight boson/scalar field regime (10⁻²⁰ to 10⁻¹⁰ eV), where few other observational techniques have constraining power. It also offers sensitivity to dark matter's spatial distribution around black holes — information that other methods cannot directly access.

The Assumptions That Could Undermine the Approach

Both studies rest on theoretical assumptions that are actively debated within the astrophysics community.

The central assumption is that dark matter forms dense concentrations — "spikes" — around black holes. The theoretical basis for this comes from adiabatic growth models: as a black hole slowly accretes mass at a galactic center, the surrounding dark matter is gravitationally compressed into a steep density profile, potentially reaching densities orders of magnitude above the galactic average [3][4].

But whether these spikes survive in realistic astrophysical environments is uncertain. Several processes could destroy or weaken them:

• Galaxy mergers and dynamical heating can redistribute dark matter, converting steep "cuspy" profiles into shallower "cored" profiles [12]. If the dark matter distribution near a merging binary is cored rather than spiked, the gravitational-wave imprint would be far weaker — potentially undetectable even with next-generation instruments.
• Baryonic feedback from star formation and supernova explosions can redistribute matter in galactic centers, further eroding dark matter spikes [12].
• Self-interactions among dark matter particles, if present, would smooth out density concentrations. One study notes that "repulsive self-interactions smoothen the 'spike' of an isolated black hole and saturate the density" [12].

If the density profiles near merging black holes are significantly lower than the spike models predict, the entire detection strategy loses sensitivity. Aurrekoetxea acknowledges part of this concern: "We could be detecting black hole mergers in dark matter environments, but systematically classifying them as occurring in vacuum" [2] — suggesting that the effect might be present but too subtle to identify with current methods.

The Case Against: Astrophysical Noise and Waveform Uncertainties

The strongest critique of extracting dark matter signals from gravitational-wave data centers on the problem of astrophysical degeneracies — other physical effects that could mimic or mask a dark matter signature.

Waveform modeling is a known challenge. Current gravitational-wave templates are semi-analytic approximations calibrated to numerical relativity simulations, and the gap between these models and true waveforms introduces systematic biases in inferred binary parameters [13]. For next-generation observatories with much higher SNR, "systematic uncertainties from calibration and waveform modeling seem likely to dominate the error budget," according to a 2023 study in Classical and Quantum Gravity [13].

Several astrophysical effects could produce phase shifts similar to those attributed to dark matter:

• Accretion disk dynamics around merging black holes can modify orbital evolution and gravitational-wave phase [14].
• Tidal effects from nearby stars or gas in dense galactic environments introduce additional perturbations.
• Eccentricity in binary orbits, if not properly modeled, can bias parameter estimation in ways that might be confused with environmental effects [14].

Distinguishing a dark matter signal from these confounders requires either extremely precise waveform models — which do not yet exist — or statistical methods applied across large populations of events. The MIT team's approach of screening many events is a step in this direction, but with only one suggestive signal out of 28, the statistical foundation remains thin.

Follow the Money: Funding and Institutional Incentives

The MIT-led study was funded by the U.S. National Science Foundation and MIT's Center for Theoretical Physics [2]. The Amsterdam study was produced at the GRAPPA center, which is supported by European research funding. None of the authors appear to have direct financial interests in detector construction.

However, the broader context matters. The Einstein Telescope, currently estimated to cost between €2.2 and €2.9 billion over nine years, is competing for a construction site decision expected in 2026–27 [15][16]. The Netherlands has reserved €870 million, Flanders €500 million, and Wallonia €200 million — already over €1.5 billion in committed funds [16][17]. Cosmic Explorer, the U.S. counterpart with planned 40-kilometer arms, has no finalized budget but would likely cost in a comparable range [9].

Source: Various sources

Data as of Dec 1, 2025CSV

Dark matter science is one of several headline justifications for these instruments. The Einstein Telescope's own documentation lists understanding "the nature of dark matter (such as primordial BHs, axion clouds, dark matter accreting on compact objects)" as a key science goal [8]. Research demonstrating that gravitational waves can probe dark matter strengthens the case for funding these observatories — creating an incentive structure that, while not evidence of bias, warrants transparency.

The LZ experiment, by contrast, cost approximately $55 million — roughly 50 times less than the Einstein Telescope's estimated price tag [10]. The entire portfolio of U.S.-funded direct-detection dark matter experiments, including SuperCDMS and ADMX-Gen2, operates on budgets in the tens of millions [10].

A Realistic Timeline

From the current theoretical proposals to a credible dark matter detection claim, the path is long:

Near term (2026–2028): LIGO-Virgo-KAGRA's next observing run, expected to begin in late 2026, will provide additional merger events for the MIT team's screening method. But with current detector sensitivity, any finding would likely remain at the "intriguing hint" level rather than a definitive detection.

Medium term (2028–2035): Detector upgrades (A# for LIGO, Advanced Virgo+) will incrementally improve sensitivity. Construction of the Einstein Telescope is expected to begin around 2028, with LISA scheduled for launch in 2035 [8][15].

Long term (2035–2040s): The Einstein Telescope and Cosmic Explorer, once operational, would provide the SNR needed to distinguish dark matter effects from astrophysical noise across large event catalogs. LISA's EMRI observations would enable the Amsterdam group's relativistic framework to be applied to real data [4][9].

This timeline — roughly 10 to 20 years from proposal to potential detection — is consistent with the pace of progress in gravitational-wave astronomy. But it also mirrors the timelines of previous dark matter detection approaches that have not yet delivered. The WIMP search program, launched in earnest in the 1990s, has progressively improved sensitivity by orders of magnitude without a confirmed signal [5]. Axion searches like ADMX have been running since the late 1990s and continue to narrow the parameter space without a detection [10].

The Research Landscape

The intersection of dark matter and gravitational-wave research is a growing field. More than 50,000 papers have been published on topics spanning both subjects, with annual output peaking at over 7,700 papers in 2023 before declining modestly to about 5,500 in 2026 [18].

Source: OpenAlex

Data as of Jan 1, 2026CSV

The LZ experiment alone involves 250 scientists from 37 institutions across multiple countries [5]. The LIGO Scientific Collaboration encompasses over 1,500 researchers. If gravitational waves were to confirm or rule out dark matter in the ultralight regime, the implications would ripple across both communities — potentially redirecting research priorities and funding away from approaches that target other mass ranges.

Global spending on dark matter experiments is difficult to precisely quantify, but U.S. federal investment alone was approximately $58 million annually as of the last major funding cycle, with comparable contributions from European and Asian partners [10]. A confirmed detection by any method would reshape these priorities; a definitive null result from gravitational-wave searches in the ultralight regime would narrow the field further but would not threaten the core programs targeting WIMPs and axions at higher masses.

What This Means

The gravitational-wave approach to dark matter detection is neither hype nor a sure bet. It is a methodologically sound expansion of the search into a mass regime that other experiments cannot access. The MIT team's finding regarding GW190728 is a proof of concept, not a discovery. The Amsterdam group's relativistic framework is a theoretical tool awaiting observatories that will not exist for another decade.

The strongest version of the case for this approach: if dark matter consists of ultralight scalar particles, and if those particles form density enhancements near black holes, then gravitational waves are among the very few observational channels capable of detecting them. That is a genuine scientific contribution.

The strongest version of the case against: the method depends on uncertain dark matter density profiles, faces formidable astrophysical noise challenges, and requires instruments that cost billions of euros and will not be operational until the 2030s. If the theoretical assumptions about dark matter spikes are wrong — as some simulations of cored halos suggest — the method may produce nothing but upper limits.

Both cases deserve to be weighed honestly as the physics community decides how to allocate the next generation of research funding.