Zombie has a great summary of the photographic frauds appearing in Reuters and other mainstream news outlets during the latest war between Israel and Hezbollah. Patterico covers the fraud coverage in the mainstream media.
Reuters reacted with admirable integrity to the first case of fraud, but it seemed that when they started to see more and more accusations of journalistic malpractice on their pages, they circled the wagons and began issuing blanket denials. That's the advantage of owning the big megaphone: if you don't admit something, then most of the world will never even hear about it.
One thing that has helped Reuters to justify ignoring all the subsequent charges is that there have been some false alarms --alleged fraud that turns out not to be well substantiated (see here and here). The false allegations help to conceal the true allegations because people will reason that if some of the allegations are false, then all of them might be, but that is because they do not understand what is really happening on the blogs. Those false allegations are not analogous to false news reports from traditional publishers, rather they are analogous to dead-end research papers in academic journals. A given author's blog serves to publish everything that the author wants to publish; they aren't, as a rule, specialized in function as traditional print media are. There aren't many pure research blogs vs. pure news blogs, for example. A given blog will post something that counts as news one hour, and the next hour post some first-impression analysis of a photograph. This analysis post is not meant to be the final word, but rather is an invitation to criticism and correction from other bloggers just like a research paper would be. Usually, a lot of the criticism appears in the comments of the post itself and the original author will often extend, modify, or retract his thesis as a result of this criticism, but even if he doesn't, you can read the comments yourself.
This process is not a failure of the blogosphere --it is an achievement; this is open-source journalism, the army of Davids at work in a process very much like science or other forms of academic research, vastly accelerated through the magic of the internet. In just a few days, this process postulated many examples of journalistic photo fraud, examined the claims, threw out the weak ones, and confirmed the many good ones. That process would have taken months or years by traditional methods of research.
So how do you know the difference between a research blog post and a news blog post? Typically, there is no way to tell except from the content, and that takes some experience to recognize. Maybe blogs should implement a tradition to help the reader differentiate between research, news, and other categories. I sometimes put the word "speculation alert" in my post headers. "Speculation" isn't research, it means that I'm just guessing about something on weak evidence, something that can't be proven one way or the other; it just seems to fit the facts (I occasionally use the even weaker conspiracy-theory alert). The tag "research alert" could be used similarly to indicated something that one is actively working on, something that is worth discussing and developing. It would clue in the reader that the blogger is not intending to make a definitive statement.
More generally, bloggers should always be aware that some of their readers don't know how blogs work, and should write with that in mind.