Product1 publisher2 min readPublished
Six of nine outside advisers quit Big Tech's counterterror group over a Meta-led plan to limit their scrutiny
Six of nine independent experts advising GIFCT, the tech industry's counterterrorism consortium, resigned Monday over a Meta-led plan to curb their scrutiny. GIFCT says the plan is not final. As drafted, the committee could still brief companies on violent trends but could not judge whether their efforts work.
The Product Desk · Product desk
What happened
- The dispute traces to a July email from Nell McCarthy, Meta's vice president overseeing content policy, saying Meta and other leaders wanted to "refresh" the six-year-old committee.
- Under the July plan, tech companies would pick new advisers instead of sitting members electing them, and the committee could not assess GIFCT's performance or make recommendations as a group.
- The resigning members wrote that their appeals against the plan were "ignored" and that they had lost confidence in GIFCT's ability to deliver on its founding mission.
- Meta passed questions to GIFCT, whose unsigned statement said the changes were "informed by several rounds of feedback."
Compiled by The Product DeskSomething wrong?How this is made
Why it matters
- constraint If companies choose the advisers and the committee cannot issue joint recommendations, no body inside GIFCT would be both independent of the members and allowed to say whether their takedown work succeeds.
- decision Meta's chair role gives it outsized sway, but Microsoft, YouTube and the other board companies still have to approve a plan that two-thirds of the independent committee resigned over.
- contradiction GIFCT says the overhaul will draw more substantive input from civil society, yet civil society representatives are among the experts who resigned over it.
GIFCT's small staff sends alerts about violent content to trust and safety leads at its roughly 35 member companies [9][10]. The same staff helps members trade threat intelligence and commissions research on extremism [10]. GIFCT also oversees a shared tip database for coordinating takedowns, and a WIRED investigation in 2024 found failures in it [11].
The advisers saw their committee as the watchdog on that work [5]. Its members were unpaid, and they elected their own successors [11][5].
GIFCT pitches the overhaul as a way to widen outside input. Its statement says the changes came out of discussions on "how to more effectively engage civil society and governments for substantive input," because "multi-stakeholderism is a core principle" [8]. The plan lets the companies choose who advises them and keeps those advisers away from the question of results [5].
The departing members had agreed with McCarthy that the consortium needed changes. They call this proposal a step backward [13]. "We all know that a body that cannot scrutinise, take a position, or evaluate is not an advisory body at all," their letter stated. "It is decoration and accountability theatre." [7] One of the departing experts told WIRED: "There won't be critical voices raising concerns about what GIFCT is doing or is not doing." [14]
You can place any advisory body on a two-by-two grid. One axis is who seats new members: the sitting members or the company being advised. The other is remit. Either the body may judge, as a group, whether the program works, or it may only brief on the problem. A self-selected body with a judging remit is a watchdog. A company-selected body with a briefing remit is a briefing service. The July plan moves GIFCT's committee from the first corner to the second [5]. Either of the mixed corners would have kept one kind of independence. One is company-picked members who can still publish a joint finding. The other is self-selected members limited to threat briefings.
Three of the nine independent members remain [1]. The experts expect the plan could be finalized soon [13].
An operator deciding how much weight to give any industry body's independent council can settle most of it by asking who seats the next member and whether the council can publish a joint finding that the program is failing. The charter answers both. When the answers are "the company" and "no", the council produces briefing material, and a member platform has to judge the program's results from its own takedown records. Under the July plan, GIFCT's answers would be "the companies" and "no" [5].
What to watch
- Whether the final text leaves the committee any way to issue a group recommendation or comment on GIFCT's performance.
- Who the companies appoint to the six vacated seats, and whether the three remaining independent members stay.
- Whether GIFCT publishes any assessment of its tip-sharing database once the committee can no longer review its results.