Escape doesn't let you do something new, it lets you do it better.
A player can normally attempt a task to get free of something. You are able to spend speed/might/intellect points to improve the odds of success, and failure may mean damage.
Escape costs 2 speed points, but there is no task to attempt. The player escapes automatically with no task. There is no chance of failure (unless there is a GM intrusion coming!), and no chance of taking damage. All of this makes Escape far safer than a task.
The question explicitly contrasts Escape to defense tasks. But Escape also applies to non-combat situations. For example, a character tied up and left in a room could use the Escape ability to get out of the restraints. No kind of defense task will help you (though you could use a Speed/Might/Intellect task to attempt to escape anyway).
This is how I adjudicate abilities in Numenera. If an ability says you do something, you do it so long as it is allowed by the fiction.
This isn't all that powerful in most circumstances
In a comment, you mentioned that this seems powerful. Does it mean that a character can spend 2 speed to overcome a difficulty 9 challenge?
No. A difficulty 9 challenge is impossible. Unless you bring down the difficulty level, it may not be overcome by any means. In a logical sense this is an "unstoppable force meets immovable object" kind of problem. I would recommend leading with the fiction to adjudicate what happens.
Is this ability overly powerful for tasks that aren't impossible? The highest "possible" difficulty is 6. A fresh tier 1 character would have to be fairly invested to have decent odds at this, unless they have access to some assets to help.
On the other hand, Escape comes at a high opportunity costs. Abilities aren't plentiful, and choosing Escape means foregoing many others. It provides a significant bonus for a number of relatively uncommon situations. At higher tiers where the player is likely to have a higher effort score and speed edge, it's usefulness will be diluted.