Timeline 01 Oct 2026 Yann LeCun: the heads of OpenAI and Anthropic saying that AI can kill us all is “incredibly destructive for everybody” and “the worst marketing campaign you can possibly imagine” Media & information · Science · AI safety · fortune.com ↗ Jul 2026 FLI AI Safety Index: the best grade among 9 companies is C+ for Anthropic; in existential safety none scores above C−; leading labs have weakened their commitments to stop at “red lines” Laws & regulation · AI safety · futureoflife.org ↗ 15 Jan 2026 For Wikipedia's 25th birthday, Amazon, Meta, Microsoft, Mistral AI and Perplexity are publicly announced as Wikimedia Enterprise partners, joining Google in paid access to its data Media & information · Business & markets · Wikipedia after AI · enterprise.wikimedia.com ↗ 15 Jul 2025 More than 40 researchers from OpenAI, Google DeepMind, Anthropic, Meta and other organizations: monitoring models' chains of thought is a rare but fragile opportunity for AI safety Science · AI safety · arxiv.org ↗ 18 Feb 2025 Open Source Initiative: Llama 3.x “is still not Open Source by any stretch of the imagination”, yet Meta keeps promoting it as open source Software development · Laws & regulation · Open source · opensource.org ↗ 03 Jan 2025 Meta removes its AI character accounts on Instagram and Facebook after an outcry; the company calls them a 2023 test Media & information · Dead internet theory · nbcnews.com ↗ Mar 2024 The Stanford Internet Observatory studies 120 Facebook Pages that posted at least 50 AI-generated images each: the Feed shows them even to users who do not follow the Pages Media & information · Dead internet theory · fsi9-prod.s3.us-west-1.amazonaws.com ↗ 20 Jul 2023 Open Source Initiative: Meta has created the misunderstanding that Llama 2 is “open source” — it is not Software development · Laws & regulation · Open source · opensource.org ↗ 18 Jul 2023 Llama 2 with open weights Software development · Business & markets · Open weights · about.fb.com ↗ 26 Mar 2023 Yann LeCun: auto-regressive LLMs are “doomed”, as the error compounds exponentially with every token, and this cannot be fixed without a major redesign Science · Large language model · x.com ↗ 24 Feb 2023 Meta releases LLaMA, a language model for researchers (non-commercial license, access on application) Software development · Business & markets · Large language model · ai.facebook.com ↗ 17 Mar 2018 The Cambridge Analytica scandal: data from tens of millions of Facebook profiles used for political advertising Elections & politics · Privacy & surveillance · Microtargeting · privacyinternational.org ↗ 04 Feb 2004 Facebook launches Media & information · Mind & relationships · Social networks · historyhit.com ↗ meta
country USA founded 2004-02 models Llama company bots Meta-ExternalAgent — training → allowed (CC BY 4.0, with a link)
Meta-WebIndexer — search → allowed
Meta-ExternalFetcher — user requests → allowed