Read on Twitter
There are so many good takeaways from this discussion about open source investigations, but one that I really like is @KAlexaKoenig's anecdote about anchoring bias and the need to have a good peer review process in investigations.
Only a few fact-finding missions have dedicated investigators with experience in geolocation, social media mining, network and source analysis, digital preservation, etc. This makes peer review slightly challenging because even though there's a low barrier to entry to learning..
..these skills, you need many to do a thorough review. You need to see how someone came to a conclusion: what material did they use, is it verifiable, how did they find it, how reliable is the source, how did they get it, etc. Great peer review replicates the entire process imo.
Even where you have a reliable source, they can inadvertently introduce false information into the investigation. I've seen many sources provide images or videos with claims that turned out to be entirely inaccurate -- and sometimes you just can't verify the content.
As someone who also studies information operations, I've thought about this last part a lot. Human rights investigators are prime targets for state-backed operations (who obviously have motivation) and in some respects they have many vulnerabilities, one of which is open source..
..content, but witness testimony and submissions by other parties as well. Even the most reliable source can accidentally introduce false information into the investigation (I've seen it). And even just one mistake, in this context, can provide fodder for even the crudest IO.
(Forgot to tag @ProfYvo and @Daragh_Murray, who were also part of this discussion and have organized a great workshop this month!) https://twitter.com/Swansea_Law/status/1356291247376392193
