A Strange New Form of Dark Matter Could Reveal Gravity’s Secrets — and Explain Three Cosmic Puzzles

For centuries, astronomers have inferred the presence of an invisible substance—dark matter—that shapes galaxies and the large-scale structure of the universe through its gravity. Yet the particles that make up dark matter remain undetected. A new proposal, published in Physical Review Letters, challenges a common assumption and suggests dark matter may sometimes collide and interact with itself.

What Is Self-Interacting Dark Matter?

Most models treat dark matter as “collisionless”: particles pass through one another and ordinary matter with negligible direct interaction, producing only gravitational effects. The new idea—self-interacting dark matter (SIDM)—posits that dark-matter particles can exchange energy and momentum when they meet, producing different structures than collisionless dark matter would.

“Imagine a crowd that deliberately bumps into one another rather than moving past each other quietly,”

— Hai-Bo Yu, University of California, Riverside, a co-author of the study.

According to Yu and colleagues, such collisions could free enough energy to create very dense, compact dark objects—cores or clumps of dark matter that are far more concentrated than typical cold, collisionless dark-matter halos. Those compact objects would be difficult or impossible to produce and detect in current particle colliders, but their gravitational fingerprints might be visible in astronomical data.

Three Astrophysical Hints That Could Fit SIDM

1. The GD-1 Stellar Stream
GD-1 is a long, thin stream of stars in the Milky Way halo that shows a prominent gap or “scar” in its density. In a 2025 Astrophysical Journal Letters paper, Yu and collaborators argued that the disturbance may require an encounter with an extremely dense object—potentially a compact SIDM clump—because conventional cold dark matter does not easily produce gaps of that size. (Note: GD-1 lies much farther than a few dozen light-years; its distance from the Sun is on the order of tens of thousands of light-years.)

2. A Peculiar Gravitational Lens (JVAS B1938+666)
Gravitational lenses—foreground masses that bend and magnify background light—can reveal small-scale structure along the line of sight. Yu suggests that the lens system JVAS B1938+666 could be influenced by an ultra-dense dark object. He and others caution, however, that complex lens systems may be affected by multiple visible and invisible components, making firm interpretation challenging.

3. The Fornax Dwarf And Fornax 6
The Fornax dwarf galaxy, discovered by Harlow Shapley, hosts several star clusters, including the oddly shaped and faint Fornax 6. Some astronomers attribute Fornax 6’s irregularity to tidal forces from its host galaxy. Yu offers an alternative: a concentrated clump of dark matter could be trapping stars and helping maintain the cluster’s structure.

Challenges And Next Steps

Experts like Yonatan Khan (University of Toronto), who did not participate in the study, emphasize caution. Key properties of dark matter—its mass, whether it couples to ordinary matter beyond gravity, and whether it interacts with itself—remain unknown. The main obstacle for proving SIDM is finding astrophysical observables that uniquely point to self-interactions rather than to ordinary astrophysical processes.

Fortunately, the coming decade will bring far deeper, wider surveys. Instruments such as the Vera C. Rubin Observatory (a ground-based wide-field survey telescope in Chile) will map stellar streams and faint structures across large swathes of sky and could provide the statistical samples needed to test SIDM predictions. Additional high-resolution lensing studies and detailed observations of dwarf galaxies will also be crucial.

Yu and Khan agree that current results are intriguing but far from definitive. If future observations confirm compact dark clumps where collisionless models cannot account for them, SIDM would offer a major new window into the particle physics of the dark sector and the workings of gravity on small scales.

Bottom Line: Self-interacting dark matter is a plausible and testable alternative to strictly collisionless models. Upcoming surveys and careful modeling may soon decide whether SIDM explains particular anomalies—or whether conventional explanations hold.