With more than 5300 exoplanets detected so far, it is clear that planet formation is a robust and efficient process. The current population of known exoplanets exhibits a wide diversity, both in nature (mass, radius) and in architecture: while giant planets can be found at large separations, the most common type of exoplanetary systems revealed by Kepler transits consist of chains of low-mass planets, super-Earths and mini-Neptunes, located close to their host stars. To understand the origin of this diversity, we need to explore the birth environment of the planets, namely the planet-forming protoplanetary disks, and to investigate their structure and evolution on both local and global scales. While considerable progress has recently been made in probing the disks on large scales (a few tens of astronomical units, au), little is known about the innermost regions (less than a few au). The IRYSS (Innermost Regions of Young Stellar Systems) project aims at deciphering the processes at play in the innermost regions of protoplanetary disks (PPDs). For the first time, we will provide a statistical view of the inner parts of a large sample of PPDs, thus bringing to light the main missing piece in our understanding of planet formation. The project builds on the unique synergy between the observational approaches developed by the partners, IPAG and IRAP, on national Research Infrastructures such as ESO/VLTI (with the PIONIER and GRAVITY interferometric instruments, largely developed at IPAG) and CFHT (with the ESPaDOnS and SPIRou spectropolarimeters, both developed at IRAP), in combination with the development of advanced physical models of the inner disk edge and of the accretion flows onto the central star. Benefiting from these world-class facilities, which are at the heart of the orientations of the call, we will conduct a multi-wavelength, multi-technique, and multi-scale investigation of the inner disk regions in a few tens of young stellar systems. We will explore the initial and environmental conditions that prevail at the time of planet formation by addressing three intrinsically interconnected pillars: 1) the morphological (asymmetry, vortex, dead zone) and physical (temperature, density) properties of the innermost scales of the protoplanetary disk, by spatially resolving at the sub-au level the near- and mid-infrared continuum emission with interferometry; 2) the magnetic star-disk interaction region, extending over a few stellar radii, and whose outer edge is thought to be the place where inwards migrating planets pile up, with spectropolarimetric observations and Zeeman-Doppler Imaging to derive the magnetic field topology and strength; 3) the dynamical timescales of the physical processes from a few days to months, by monitoring the variability of both the magnetic topology and the small-scale disk features. The combined analysis of these data sets arising from these two state-of-the-art observational techniques will put the world-leading French experts in a unique position to provide the stellar and exoplanet communities with legacy databases of magnetic maps, line profiles, inner rim positions and disk substructures. These are the key ingredients to relate the magnetic properties of young stars to the structure of their inner disk, and to investigate their evolution over periods as long as 10 years for some emblematic objects. As such, this legacy will provide access to a detailed overview of the innermost regions of nascent stellar systems and their disks where close-in planets form. Our team has access to all the ESO and CFHT Large Program and Guarantee Time observations to be exploited in the IRYSS project, and has developed during previous ERC-funded grants cutting-edge analysis and modeling tools required for their interpretation. We therefore gather the optimal expertise to yield major advances in this competitive field, supported by the appropriate workforce provided by this specific and quite timely ANR call.