Automated tow placement (ATP) process-induced defects such as wrinkles and tow folding limit its applicability for curved paths and complex geometries and can degrade the strength of cured composite laminates. Tow tackiness or in situ bond strength-toughness (IBST) is a key property associated with tow-tow adhesion and defect formation during ATP processing. Since prior research indicates that wrinkle defect formation during ATP occurs under locally mixed-mode loading conditions, it is important to characterize mixed-mode decohesion response. To do so, this study investigates an experimental technique using a modified Arcan fixture to determine the mixed-mode decohesion response of uncured IM7-G/8552 carbon fiber/epoxy tows in terms of fracture mechanics-based traction–separation relationship. This fixture comprises two parts with holes at the outer ends, facilitating easy adjustment of loading orientation for testing under different modes using a single setup. Tow specimens are prepared under a selected temperature, pressure, and contact times, featuring a pre-crack introduced via a PTFE film. The prepared specimen is then bonded to rigid steel beams in the Arcan fixture using a cyanoacrylate rapid-curing adhesive. High-speed DIC displacement measurements of the beams enable extraction of the traction–separation relationship for the uncured tow-tow interface which is used to analyze the effect of mode mixity on the tack response. While tow-tow cohesion is achieved at a contact pressure of 0.23 MPa, contact time of 1 s, and a bonding temperature of 40 °C, decohesion experiments are performed at a debonding rate of 5 mm/s under 0°—mode-I and 30° mixed-mode conditions. Under these bonding conditions, the peak tractions in the 30° mixed mode case are nearly six times greater, and the energy required for debonding is 4.5 times higher compared to the 0° mode-I case.