# Top ~60 Influential AI Voices on X — Mid-2026 (research agent report, 2026-07-17)

Curated for a daily-signal feed: weighted toward accounts that **actively post**, with affiliations verified against mid-2026 reporting (Karpathy → Anthropic May 2026; LeCun → AMI Labs; Murati → Thinking Machines; Sutskever → SSI CEO).

## 1. Frontier-Lab Leaders

| Name | X Handle | Affiliation / Role | Why they matter |
|---|---|---|---|
| Sam Altman | @sama | CEO, OpenAI | Highest-reach lab account; product drops, AGI-timeline takes, and posts that move markets. |
| Greg Brockman | @gdb | President/co-founder, OpenAI | More technical than Altman — training-infra details, live demos, frequent poster. |
| Dario Amodei | @DarioAmodei | CEO, Anthropic | Posts less often, but essays and scaling/safety framing set the discourse for weeks. |
| Demis Hassabis | @demishassabis | CEO, Google DeepMind; Nobel laureate | Measured updates on Gemini, AlphaFold-lineage science, and AGI capability debates. |
| Elon Musk | @elonmusk | Founder, xAI | Noisy feed, but Grok releases and xAI strategy land here first. |
| Mira Murati | @miramurati | Founder/CEO, Thinking Machines Lab (fmr OpenAI CTO) | Strategic/product perspective from the most-watched new lab. |
| Ilya Sutskever | @ilyasut | CEO, Safe Superintelligence Inc. | Rare poster, but every post is an event; declared 2026 the start of a "new research era." |
| Mustafa Suleyman | @mustafasuleyman | CEO, Microsoft AI; DeepMind co-founder | Copilot direction plus policy/risk commentary from inside the largest distributor of AI. |
| Yann LeCun | @ylecun | Exec Chair, AMI Labs (left Meta Nov 2025) | The loudest "LLMs are a dead end" counterweight; now backing world models with $1B+ seed. |
| Alexandr Wang | @alexandr_wang | Chief AI Officer, Meta Superintelligence Labs | Runs Meta's frontier push post-LeCun; signals on Llama-successor direction. |
| Arthur Mensch | @arthurmensch | CEO, Mistral | The European sovereign-AI position; open-weight release announcements. |
| Clem Delangue | @ClementDelangue | CEO, Hugging Face | Barometer of the open-source ecosystem; posts daily on models/datasets/trends. |
| Aravind Srinivas | @AravSrinivas | CEO, Perplexity | Prolific poster on AI search, browser wars, product strategy. |
| Sundar Pichai | @sundarpichai | CEO, Google | Lower volume, but Gemini-era strategic announcements originate here. |

## 2. Prominent Researchers

| Name | X Handle | Affiliation / Role | Why they matter |
|---|---|---|---|
| Andrej Karpathy | @karpathy | Anthropic (pretraining research lead, joined May 2026; fmr OpenAI/Tesla/Eureka Labs) | The single most-read technical explainer in AI; his moves and threads reset field consensus. |
| Jeff Dean | @JeffDean | Chief Scientist, Google DeepMind | Systems/scaling/TPU-infrastructure signal from the person who built Google's stack. |
| Noam Brown | @polynoamial | Research scientist, OpenAI | Led o1/reasoning-model line; required reading on self-play, RL, and test-time compute. |
| François Chollet | @fchollet | Ndea co-founder; ARC Prize | The rigorous skeptic — benchmark critique and abstraction/reasoning research. |
| Jim Fan | @DrJimFan | Sr. Research Manager, NVIDIA | Best threads on robotics, embodied AI, and foundation models for the physical world. |
| Lilian Weng | @lilianweng | Research scientist (fmr OpenAI safety lead; 2026 lists place her at DeepMind) | Her RLHF/agents blog posts are canonical; X is where they preview. |
| Sebastian Raschka | @rasbt | Author/researcher (*Build an LLM From Scratch*) | Hands-on architecture breakdowns of every new model within days of release. |
| Fei-Fei Li | @drfeifei | Stanford HAI; World Labs | Spatial intelligence research plus AI-policy voice with institutional weight. |
| Andrew Ng | @AndrewYNg | DeepLearning.AI / AI Fund | Calm, reliable applied-ML signal; large-audience educational framing. |
| Chris Olah | @ch402 | Co-founder, Anthropic (interpretability) | Mechanistic interpretability's founding voice; posts when interp results matter. |
| Neel Nanda | @NeelNanda5 | Google DeepMind (mech interp lead) | Most active interpretability poster; open research + hiring/field meta. |
| Jan Leike | @janleike | Alignment lead, Anthropic | Frontline alignment-research updates and candid safety takes. |
| Amanda Askell | @AmandaAskell | Anthropic (Claude character/behavior) | The philosophy of how models should behave — highly cited in every persona debate. |
| Sholto Douglas | @_sholtodouglas | Anthropic (RL scaling) | High-signal on RL scaling and capability trajectories; frequent podcast presence. |
| David Ha | @hardmaru | Co-founder, Sakana AI | Nature-inspired/evolutionary AI research plus the best paper-aesthetic curation feed. |

## 3. Engineers / Builders

| Name | X Handle | Affiliation / Role | Why they matter |
|---|---|---|---|
| Simon Willison | @simonw | Datasette/`llm` CLI; Django co-creator | The authority on what LLMs can *actually* do — prompt injection, tooling, model reviews. |
| Boris Cherny | @bcherny | Anthropic — creator of Claude Code | Agentic-coding workflows and feature previews from the tool's author. |
| Logan Kilpatrick | @OfficialLoganK | Google DeepMind — Gemini API / AI Studio lead | The most useful account for Gemini API changes and dev-facing launches. |
| Michael Truell | @mntruell | CEO/co-founder, Cursor | AI-native IDE category leader; product and model-partnership signals. |
| Lee Robinson | @leerob | VP Developer Experience, Cursor (fmr Vercel) | High-volume, practical AI-coding content with big builder reach. |
| Amjad Masad | @amjadmasad | CEO, Replit | Agents-for-everyone thesis; frequent, opinionated posts on autonomous coding. |
| Guillermo Rauch | @rauchg | CEO, Vercel | AI SDK / v0 ecosystem; where web dev meets AI tooling. |
| Pieter Levels | @levelsio | Indie hacker | Ships and monetizes AI products solo in public; the build-in-public benchmark. |
| Peter Steinberger | @steipete | OpenAI (2026); creator of OpenClaw | Viral agentic-tooling builder; now inside OpenAI — worth tracking closely. |
| Mckay Wrigley | @mckaywrigley | Takeoff AI | Daily practical posts on shipping with AI coding agents. |
| Riley Goodside | @goodside | Staff Prompt Engineer, Scale AI | Inventor of the modern prompt-engineering demo; finds model quirks first. |
| Eric Zakariasson | @ericzakariasson | Engineer, Cursor | Documents AI-native development UX experiments from inside the leading tool. |

## 4. Analysts / Commentators / Curators

| Name | X Handle | Affiliation / Role | Why they matter |
|---|---|---|---|
| swyx (Shawn Wang) | @swyx | Latent Space podcast/newsletter | Defined "AI engineer" as a category; maps the ecosystem better than anyone. |
| Ethan Mollick | @emollick | Professor, Wharton | The reference account for what AI actually does to knowledge work; posts experiments daily. |
| Zvi Mowshowitz | @TheZvi | Don't Worry About the Vase | The most comprehensive weekly AI digest in existence; misses nothing. |
| Dwarkesh Patel | @dwarkesh_sp | Dwarkesh Podcast | Best long-form AI interviews; clips and takes drive weeks of discourse. |
| Nathan Lambert | @natolambert | Ai2; Interconnects newsletter | Best curator on RLHF, post-training, and the open-weight ecosystem. |
| Jack Clark | @jackclarkSF | Co-founder, Anthropic; Import AI | Measured research+policy synthesis with a government lens. |
| AK | @_akhaliq | Paper curator (Hugging Face) | The default feed for new arXiv papers and model releases, within hours. |
| Rowan Cheung | @rowancheung | The Rundown AI | Highest-volume mainstream AI news; surfaces under-the-radar launches. |
| Nathan Benaich | @nathanbenaich | Air Street Capital; State of AI Report | Investor-grade analysis of compute, geopolitics, and the AI economy. |
| Nathan Labenz | @labenz | Cognitive Revolution podcast | Thoughtful capability-and-alignment threads; "AI scouting" for practitioners. |
| Deedy Das | @deedydas | Menlo Ventures | One-tweet distillations of model releases, comp data, and industry moves. |
| Gary Marcus | @GaryMarcus | Scientist/author | The persistent bear case; useful friction against hype cycles. |
| Eliezer Yudkowsky | @ESYudkowsky | MIRI | The originating voice of AI x-risk; still frames the doom side of the debate. |
| Lex Fridman | @lexfridman | Podcast host | Long-form interviews where lab leaders say things they don't post. |
| Allie K. Miller | @alliekmiller | AI advisor/investor | Largest business-AI audience; how enterprise hears about AI. |

## 5. High-Signal, Lower-Fame (what practitioners actually read)

| Name | X Handle | Affiliation / Role | Why they matter |
|---|---|---|---|
| Teortaxes | @teortaxesTex | Independent analyst | The DeepSeek/China open-model watcher; earliest credible reads on Chinese frontier labs. |
| Tibor Blaho | @btibor91 | Engineer, The AI Advantage | Digs unreleased features out of app builds; scoops official announcements reliably. |
| Cameron Wolfe | @cwolferesearch | Deep (Learning) Focus newsletter | Rigorous long-form paper breakdowns that practitioners bookmark. |
| wh | @nrehiew_ | Independent | Dense architecture/training-paper threads; frequently cited by lab researchers. |
| vik | @vikhyatk | Moondream | Practical small-VLM engineering; antidote to frontier-only thinking. |
| Linus Lee | @thesephist | Notion (AI research) | LLM internals, embeddings, and AI-native interface design. |
| Vaibhav (VB) Srivastav | @reach_vb | Hugging Face | Fastest signal on new open-weight releases and how to run them. |
| Omar Sanseviero | @osanseviero | Google DeepMind dev rel | Open-model ecosystem + Gemma releases; consistent hands-on content. |
| Maxime Labonne | @maximelabonne | Liquid AI | Fine-tuning/merging guides the OSS community actually uses. |
| Teknium | @Teknium1 | Co-founder, Nous Research | Open post-training frontier; Hermes models and dataset craft. |
| kalomaze | @kalomaze | Independent → post-training research | Sampling and RL tinkering in public; where inference tricks surface first. |
| xjdr | @_xjdr | Independent (entropix) | Inference-time sampling research that repeatedly jumps into lab practice. |
| Ahmad Osman | @TheAhmadOsman | Systems/infra engineer | GPUs, inference perf, and self-hosted LLM setups — the metal underneath. |
| Artificial Analysis | @ArtificialAnlys | Independent benchmark org | Neutral price/latency/quality benchmarks every release day. |
| Epoch AI | @EpochAIResearch | Research org | Compute, data, and scaling-trend analysis that anchors capability forecasts. |

### Notes on freshness (mid-2026)
- **Karpathy** joined Anthropic's pretraining team May 19, 2026 (under Nick Joseph) — still the field's top educator-voice.
- **LeCun** left Meta Nov 2025; AMI Labs raised a $1.03B seed (Mar 2026) for world models.
- **Sutskever** became SSI CEO after Daniel Gross left for Meta; SSI at ~$32B valuation, no product.
- Dario Amodei and Ilya Sutskever post infrequently — kept because their rare posts outweigh most daily feeds; everyone else listed is an active poster.
- Deliberately excluded: pure hype/affiliate "10 AI tools" accounts, and rumor accounts (e.g. Jimmy Apples) whose reliability is contested.

### Sources
- https://pasqualepillitteri.it/en/news/3633/ai-x-twitter-accounts-to-follow-2026
- https://www.softwarethug.com/posts/top-ai-people-to-follow-on-x-twitter/
- https://www.zarifautomates.com/blog/best-ai-twitter-x-accounts-to-follow
- https://amperly.com/best-artificial-intelligence-twitter-accounts/
- https://www.kdnuggets.com/10-best-x-twitter-accounts-to-follow-for-llm-updates
- https://tweetstorm.ai/blog/top-ai-influencers
- https://x.feedspot.com/artificial_intelligence_twitter_influencers/
- https://techcrunch.com/2026/05/19/openai-co-founder-andrej-karpathy-joins-anthropics-pre-training-team/
- https://www.cnbc.com/2026/05/19/anthropic-hires-openai-cofounder-andrej-karpathy-former-tesla-ai-lead.html
- https://the-decoder.com/you-certainly-dont-tell-a-researcher-like-me-what-to-do-says-lecun-as-he-exits-meta-for-his-own-startup/
- https://www.buildfastwithai.com/blogs/yann-lecun-ami-labs-world-models
- https://www.eweek.com/news/ilya-sutskever-ceo-safe-superintelligence/
- https://www.startuphub.ai/ai-news/ai-figures/2026/figure-ilya-sutskever-ssi-financial-breakdown-2026-06-06
- https://www.hyperwriteai.com/blog/top-ai-influencers-to-follow-on-twitter