Tanks, enchanters, and some fighters having all-encompassing defensive measures is a big contributor to current damage creep, but I'm mostly going to be talking about tanks. Because they block and reduce **all** damage they force interactions with them into a stat check and this effect is felt the hardest in 1 on 1 laning encounters - usually against fighters. We see this happen again and again with tanks: Nautilus top back in S5/S6, Ornn top, Sion, Shen, Poppy, etc. force the lane into an interaction of "get through my shield or you lose the trade by default." Sejuani also does this, though she is at least vulnerable to being poked back and denied her steroid. This is also part of why tank metas are so insufferable; when tanks are in a position to get fed, they can completely take over the entire game because *nobody* can handle their high, generic durability. Damage reduction has the same issue, which is why Alistar is generally a recurring problem for being able to outright ignore all incoming damage except for true - I doubt it's coincidence that we now have more true damage. (Though to be fair damage reduction is more often found on fighters.)
Because of this interaction they tend to have disproportionately strong lanes into fighters they're not supposed to be actively out-trading and beating given their natural ability to out-scale. As a response, there's been an ongoing arms race between damage and durability that is currently heavily weighted toward the former. The solution to this eternal powering up between the two isn't to also buff defense further, but to revise it away from generalist "counter all damage" designs and focusing tank survivability into being especially durable against specific things. Through this, generic solutions like giving everyone absurd amounts of armor pen or just outright ignoring defense entirely (Conqueror) can be safely taken out of the game once they're no longer a necessity for fighters to exist.
There are some tanks who do not have this issue and instead have specialized defenses. Malphite has the issue to a smaller degree by being heavily incentivized to prioritize armor, making his shield (and himself) more vulnerable to magic damage, though in his rework he should likely lose the shield altogether. Galio's magic shield doesn't protect him against physical assault at all, and his damage reduction is halved against physical damage. Amumu passively reduces AA damage on his E as well as reducing its cooldown when he's right clicked -- as well as disarming on his ult, a combination that leaves him vulnerable to casters. Things like this are much better for the game, IMO - which is why 2 of them currently struggle to justify their place, because any other tank can either just pop out massive shields, hefty heals, or just gaining tons of resists, and because the game has amped damage and tank-killing in response to that, Amumu and Malphite have found themselves on the short end of the stick.
Related tangent is that IMO tank designs are gimped in this regard by Riot's strange refusal to give tanks interesting buffs and debuffs. Ornn was _close_ to this with Brittle but the purpose of the debuff is ruined by the fact that he can proc it himself and that it's mostly just there as a conditional damage burst. Tanks who could bolster allies in specific ways or who had defenses that worked to target different means of damage (for example, countering AD spells vs AD rightclicks vs on-hit rightclicks) would immediately have ways to justify themselves being picked beyond simply being "a good tank," which helps curb the current soloQ issue of "if you didn't pick Sion why did you pick a tank at all?" Not every tank should do these things, obviously, but the current tank design of being good almost entirely for lock-down and generic bulkiness and not much else leads to a very repetitive process of only picking whichever tank has the best or most reliable CC and tankiness. (Sion in SoloQ and Sejuani in pro play for solo tanks, Alistar for support)