See other failures under 000.html.
Ritchey carbon seatpost breaks at head. From https://www.flickr.com/photos/63373992@N07/12069159753 as of 2014/08.
No reports on rider injuries. Service history unknown. No notes on how the failure was discovered.
Text at the page above says "Everybody know carbon fiber breaks [...]". In a little bit more detail, the issue here is not "carbon fiber" but "how carbon fiber gets used". Metals tend to have a range where "damage starts" and it takes a lot of added energy to get to where "part failed". In contrast, carbon fiber has a much lower energy from "starts" to "failed".
There are thus various ways to use carbon fiber to build parts. One way is parts "as strong" as metal, but once failure starts, they fail much sooner — that is, at much lower energy. Another is "fails at the same energy, but much stronger." Carbon fiber is lightweight, so even the stronger/same-failure part would probably be lighter than the metal part, and would probably be vastly more durable in common use.
However, in a market where "light weight is all that matters", most makers build parts much closer to "no-stronger/easier-failure". That means lower material cost and a lighter part, but also parts that fail as often as metal but more often fail suddenly and which fail leaving jagged edges — rather than, say, cracking but not snapping off.
In practice "carbon is more dangerous" is a good rule of thumb, but it does a dis-service to makers/parts which use the material more smartly — and may ultimately discourage smart-built parts, as they are somewhat heavier and possibly more expensive (due to higher material costs) than the more-dangerous parts.
See also FAIL-.html.