How We Use Artificial Intelligence at the GIOR

We’ve received sev­er­al ques­tions about the use of arti­fi­cial intel­li­gence in con­nec­tion with our recent Chat­G­PT-gen­er­at­ed report on the Sumud Flotil­la. It’s a fair question—and one that speaks to how we man­age an enor­mous body of mate­r­i­al with lim­it­ed time and resources.

The Glob­al Influ­ence Oper­a­tions Report (GIOR) relies heav­i­ly on arti­fi­cial intel­li­gence to process, sum­ma­rize, and ana­lyze a vast amount of open-source infor­ma­tion. The scope of what we cover—spanning mul­ti­ple coun­tries, orga­ni­za­tions, and languages—would be impos­si­ble to sus­tain by hand with the staffing and resources avail­able to us. AI is used chiefly in the “con­text sum­maries” that accom­pa­ny indi­vid­ual posts and in more exten­sive research reports labeled “Chat­G­PT.”

AI allows us to iden­ti­fy nar­ra­tives, map cross-bor­der con­nec­tions, and syn­the­size devel­op­ments at a pace no small human team could match. This effi­cien­cy comes with trade-offs. AI-gen­er­at­ed sum­maries and reports can intro­duce not only awk­ward phras­ing but also fac­tu­al or inter­pre­tive errors. We review and fact-check all AI-gen­er­at­ed mate­r­i­al to the extent pos­si­ble, but that review is nec­es­sar­i­ly cur­so­ry giv­en the scale of our out­put. That said, we do believe the over­whelm­ing pre­pon­der­ance of this mate­r­i­al is both accu­rate and fair­ly portrayed.

The pur­pose of this approach isn’t to replace expertise—it’s to extend it. Using AI allows us to mon­i­tor a far wider field and pro­vide time­ly insight where full man­u­al report­ing would be impos­si­ble. We accept that this some­times means inac­cu­ra­cies or styl­is­tic imper­fec­tions, but it also means the work can be time­ly, rel­e­vant, and com­pre­hen­sive (not that human gen­er­at­ed work is free from error!). Due to these lim­i­ta­tions, it is cru­cial that read­ers ver­i­fy any spe­cif­ic piece of infor­ma­tion from AI-gen­er­at­ed mate­r­i­al before rely­ing on it. This prin­ci­ple is reflect­ed in our stand­ing dis­claimer, which appears on every rel­e­vant post.

We remain trans­par­ent about our meth­ods and con­tin­ue refin­ing them as the tech­nol­o­gy evolves. AI gives us reach; human over­sight gives us purpose.

(NOTE: We are pub­lish­ing this expla­na­tion as a post rather than an announce­ment because of its impor­tance to our ongo­ing work and to read­ers’ under­stand­ing of how that work is produced.)