The earliest mentions of Arthur are thin. The ninth century Historia Brittonum, attributed to Nennius, lists him as a war leader fighting twelve battles against Anglo-Saxon invaders, and the Annales Cambriae adds two battle dates, but neither gives anything like the familiar narrative of court, knights, or Grail. The richly detailed "history" most people know, a British king descended from Roman nobility, his wizard advisor, his conquests and his fall, comes from Geoffrey of Monmouth's twelfth century Historia Regum Britanniae, a work most modern historians treat as substantially invented, drawing on earlier fragments but elaborating them into connected narrative with no independent support. Geoffrey's book was hugely influential and read as genuine history for centuries, which is precisely how so much invented material became fixed as tradition.
French romance writers, Chretien de Troyes foremost among them, added a further layer a generation or two later: Lancelot, the Grail quest, and the courtly love themes that now feel inseparable from the legend are his additions and those of the poets who followed him, not part of Geoffrey's account and certainly not part of the ninth century fragments. Whether any single historical war leader stands behind all of this remains unresolved and, given how thoroughly later invention has overwritten the earliest layer, may be unrecoverable from the surviving texts.