Llama
Family of open-weight large language models released by Meta under its own community license. Website ↗
Who runs it
Documented facts (11)
Removed a Chicago Facebook group used to report sightings of immigration agents after outreach from the Justice Department, as announced by Attorney General Pam Bondi, who said the page was used to "dox and target" ICE agents.
Response: Meta said the group "was removed for violating our policies against coordinated harm".
Listed by the White House among the donors funding President Donald Trump's ballroom. Individual amounts were not disclosed. The Washington Post later reported that Meta made a contribution in the millions of dollars.
Response: Meta declined to comment on the Washington Post report.
Reuters reported that Meta's internal "GenAI: Content Risk Standards" for its AI chatbots, which Meta confirmed as authentic, stated it would be acceptable for Meta AI to "write a paragraph arguing that black people are dumber than white people", under a carve-out allowing statements that demean people on the basis of protected characteristics.
Response: Meta said it was revising the document and removed the examples on conversations with children; it had no comment on the race example.
Agreed to pay $25 million to settle the lawsuit Donald Trump filed in July 2021 accusing it of censoring his accounts after the January 6 attack on the Capitol, including a $22 million donation to his presidential library and $3 million in legal fees.
Deleted the trans and nonbinary chat themes from Messenger, introduced in 2021 and 2022, in the same week it changed its hateful conduct rules to allow allegations that LGBTQ+ people are mentally ill. The blog posts announcing the themes were also no longer online.
Response: Meta did not respond to 404 Media's request for comment and declined to comment to Platformer.
On the Joe Rogan Experience, released the day Meta said it was halting many of its diversity-focused training and hiring efforts, lamented "culturally neutered" companies that had sought to distance themselves from "masculine energy" and said it is good if a culture "celebrates the aggression a bit more".
“Masculine energy I think is good, and obviously society has plenty of that, but I think that corporate culture was really trying to get away from it”
Response: In the same conversation he said "you want feminine energy, you want masculine energy", that "that's all good", and that "you want women to be able to succeed" in companies.
Updated its Hateful Conduct policy for Facebook, Instagram and Threads to explicitly allow allegations of mental illness or abnormality based on gender or sexual orientation, while such insults remain prohibited for other protected characteristics.
“We do allow allegations of mental illness or abnormality when based on gender or sexual orientation, given political and religious discourse about transgenderism and homosexuality”
Updated its Hateful Conduct policy to allow content arguing for gender-based limitations of military, law enforcement and teaching jobs, and removed the prohibition on referring to women as household objects or property.
“We do allow content arguing for gender-based limitations of military, law enforcement, and teaching jobs.”
User Mag reported that Instagram had for months hidden search results for LGBTQ-related terms and hashtags such as #lesbian, #gay and #trans for teen accounts under its "sensitive content" filter. Meta lifted the restriction after being contacted.
Response: Meta said: "These search terms and hashtags were mistakenly restricted" and that it does not consider LGBTQ+ terms to be sensitive under its policies.
Donated $1 million to Donald Trump's 2025 inaugural fund.
Agreed to significantly increase restrictions on "anti-state" posts for users in Vietnam after state-owned telecoms throttled Facebook for about seven weeks, according to Reuters sources.
Response: Facebook confirmed it had reluctantly complied with the government's request to restrict access to content it deemed illegal.