Usually you have to first determine if the crankarm spider ars are bent, or if it's the ring itself.

I have reliably straightened hundreds of crankarm spiders using a 3lb hammer and broom-handle dowel as a drift punch, applied to the base area of the spider tabs where the chainring appears to meet the crankarm.
It's a job for someone with experience bending metal, and I have never broken a tab off doing this.

OTOH, if you try to bend a chainring while it's mounted to the spider, there's a LOT of leverage and I HAVE broken a crank spider tab doing that.

Trueing cranks and chainrings is a skill that isn't so much fun to learn, probably why so few mech's are confident enough to have a serious go at it. It's also more profitable to sell a new part, even if the replacement chainring is a poor-shifting aftermarket part.
It has been worthwhile for me though, since it saves a lot of parts, and a true-running chainring really improves the ability to adjust the front derailer to perfection.

I wouldn't rule out some judicious force applied to a mounted chainring, but be forewarned about the amount of leverage you're playing with. I think I was using an adjustable wrench when I broke that tab, but it is also possible that it was already broken before I tried bending it. FWIW, it was a cheap, cast Suntour GPX crank.

Bending a chainring back, with it removed and on the bench, is almost impossible.
The ring part may be straight, but if the tabs are then even slightly out of parallel, the ring goes out of plane as soon as the chainring bolts are tightened. It doesn't help that the rings are cold-rolled and relatively springy.

Might be a good time to look for some newer round chainrings?