Immediately, X launched the corporate’s first transparency report since Elon Musk purchased the corporate, previously Twitter, in 2022.
Earlier than Musk’s takeover, Twitter would launch transparency studies each six months.These largely coated the identical floor as the brand new X report, giving particular numbers for takedowns, authorities requests for data, and content material removals, in addition to knowledge about which content material was reported and, in some instances, eliminated for violating insurance policies. The final transparency report accessible from Twitter coated the second half of 2021 and was 50 pages lengthy. (X’s is a shorter 15 pages, however requests from governments are additionally listed elsewhere on the corporate’s web site and have been persistently up to date to stay in compliance with varied authorities orders.)
Evaluating the 2021 report back to the present X transparency report is a bit tough, as the best way the corporate measures various things has modified. As an example, in 2021, 11.6 million accounts had been reported. Of this 11.6 million, 4.3 million had been “actioned” and 1.3 million had been suspended. In line with the brand new X report, there have been over 224 million studies, of each accounts and items of particular person content material, however the end result was 5.2 million accounts being suspended.
Whereas some numbers stay seemingly constant throughout the studies—studies of abuse and harassment are, considerably predictably, excessive—in different areas, there’s a stark distinction. As an example, within the 2021 report, accounts reported for hateful content material accounted for almost half of all studies, and 1 million of the 4.3 million accounts actioned. (The studies was interactive on the web site; the present PDF not permits customers to flip by means of the info for extra granular breakdowns.) Within the new X report, the corporate says it has taken motion on solely 2,361 accounts for posting hateful content material.
However this can be on account of the truth that X’s insurance policies have modified because it was Twitter, which Theodora Skeadas, a former member of Twitter’s public coverage crew who helped put collectively its Moderation Analysis Consortium, says may change the best way the numbers look in a transparency report. As an example, final yr the corporate modified its insurance policies on hate speech, which beforehand coated misgendering and deadnaming, and rolled again its guidelines round Covid-19 misinformation in November of 2022.
“As sure insurance policies have been modified, some content material is not violative. So in the event you’re taking a look at adjustments within the high quality of expertise, that could be onerous to seize in a transparency report,” she says.
X has additionally misplaced customers since Musk’s takeover, additional complicating what the brand new actuality of the platform may appear to be. “When you account for altering utilization, is it a decrease quantity?” she asks.
After taking on the corporate in October of 2022, Musk fired nearly all of the corporate’s belief and security employees in addition to its coverage employees, the individuals who make the platform’s guidelines and guarantee they’re enforced. Below Musk, the firm additionally started charging for its API, making it more durable for researchers and nonprofits to entry X knowledge to see what was actually occurring on the platform. This may occasionally additionally account for adjustments between the 2 studies.
