As mentioned in the comments, the definitions were probably inserted by later authors, like Heron, for didactic purposes and/or for compliance with Aristotelian ideas about proper scientific exposition (which starts with "intuitive definitions" that fix the subject matter). There is plenty of textual evidence to support this, e.g. Russo discusses it at length in The Definitions of Fundamental Geometric Entities Contained in Book I of Euclid’s Elements
"Many of the ideas here reported by Heron can be found in Aristotle’s writings. Among Aristotle’s passages on this subject we may quote, e.g., Phys., IV, 11, 220a 15 ff., where, among other things, points are said extremities of lines and the analogy between point and instant of time is introduced; De Cael., III, 1, 300a 14, where the above analogy is also considered; Met., V, 6, 1016b 24–30, where that which is indivisible and has position is called point. This definition of a point as a monad having position is in fact an ancient Pythagorean definition, as we know from Proclus (95, 22).
As another example, we may recall that the definition of lines as extremities of surfaces, which is contained in a passage of Heron’s definition 2 (corresponding to definition I, 6 of the Elements), is the Platonic definition which Aristotle had criticized in Topica, VI, 6, 143 b 11. In other instances Heron may have used later sources than Euclid: for example
the definition of a straight line, as we shall see in detail in Sect. 10, seems to be drawn from Archimedes."
Another, not necessarily incompatible, speculation about the functionality of the Book I definitions is given by Azzouni in Proof and Ontology in Euclidean Mathematics. It has to do with Euclid's essential use of diagrams, and the dissimilarity of his approach to modern axiomatic method, where such definitions are of no use because nothing is (supposed to be) inferred from the diagrams. This would explain why modern mathematicians are usually perplexed by these definitions, given their, along with the diagrams', "lack" of mathematical utility.
On the other hand, if one is concerned with which diagrams are admissible for demonstrations then having "fat" points, for example, has to be explicitly proscribed. Indeed, Epicureans objected to Euclid's demonstration of proposition I.1 on the basis that dimensionless points do not exist, the circles intersect over a segment, and the proof fails.
"If these uses for diagrams are backtracked to ancient Greek mathematics, an analogous role can perhaps be found for their definitions of primitive terms as well: They are heuristic guides that give the student an intuitive grasp of otherwise difficult mathematical notions. Support for this view can be found in the fact that such definitions play no role in subsequent proofs. Unfortunately, Greek mathematicians took these definitions more seriously than the above interpretation can allow: Their constant tinkering with the definitions for straight lines, points, and so on, suggests they did not see them as heuristic.
Indeed, the same point applies to their diagrams: The Greek focus on whether mechanical methods for constructing curves were admissible suggests to some that (some) Greek mathematicians had constructivist scruples about mathematical objects... It's possible, however, that constructivist scruples (at least as they appear in the tradition leading to book I) were concerned with what sorts of diagrams are admissible in proofs, and not with mathematical objects at all... My conjecture (which only historians can decisively confirm or not) is that we see in the first book of Euclid's Elements a blend of a “pictorial” proof-system (in which diagrams are an essential part of the proofs), and a language-based proof-system (in which diagrams are merely heuristic).
Alas, it is doubtful that historians can decisively confirm or infirm such conjectures.
