AI Safety (2026)

If you want to learn about AI Safety and AI Ethics, I recommend following a mix of researchers, philosophers, technologists, and policy experts rather than relying on one perspective.

The most important thing is to understand both sides of the field:

• AI Safety: How do we make increasingly powerful AI systems reliable, controllable, and aligned with human values?
• AI Ethics: How do we ensure AI respects human rights, fairness, privacy, accountability, and human dignity?

Here are some excellent people to start following.

1. People I strongly recommend following

AI safety and alignment

1. Yoshua Bengio

Top recommendation

One of the leading voices on advanced AI safety, existential risks, and responsible AI development.

Follow for: Scientific research, AI risk, and global safety.

2. Stuart Russell

Explains why AI systems must be designed around human values and human well-being.

Follow for: AI alignment, human-compatible AI, and ethics.

3. Neel Nanda

Makes technical AI safety research more approachable.

Follow for: Mechanistic interpretability and understanding how AI models work.

AI ethics, fairness, and human rights

4. Timnit Gebru

A major voice on algorithmic bias, representation, accountability, and the social consequences of AI.

Follow for: Ethical AI and responsible research.

5. Joy Buolamwini

Researches algorithmic discrimination and facial recognition bias.

Follow for: AI fairness, human rights, and algorithmic justice.

6. Abeba Birhane

Challenges the assumptions behind AI development and explores its effects on society.

Follow for: AI accountability, power, and social justice.

7. Shannon Vallor

Particularly valuable for understanding the philosophical dimensions of AI ethics.

Follow for: Human values, technology ethics, and moral responsibility.

AI governance and public policy

• Helen Toner – AI governance, oversight, and national security.
• Arvind Narayanan – AI accountability, misinformation, and the limitations of AI systems.
• Rumman Chowdhury – Algorithmic auditing, accountability, and responsible AI.

These voices cover different aspects of the field. People to follow in AI safetyVoices and perspectives on AIThe Who's Who of AI | AI Weekly - AI News & AI Newsletter on Artificial Intelligence & Deep LearningAI governance experts

2. Organizations worth following

These are particularly useful because they publish research, educational material, and policy discussions.

Center for Human-Compatible AI

Human-compatible AI and alignment research.

Website

Distributed AI Research Institute

AI ethics, power, labor, and social impact.

Website

AI Now Institute

AI accountability, regulation, and human rights.

Website

Future of Life Institute

Advanced AI safety and existential risk.

Website

Stanford HAI

Human-centered AI research, governance, and ethics.

Website

3. My recommended learning path

If you're starting from scratch, I'd approach it in this order:

One particularly useful starting resource is the International AI Safety Report 2026. It brings together international research on AI capabilities, risks, and safety. International AI Safety Report 2026Contributors

4. My personal recommendation for your starting five

If you don't want to overwhelm yourself, start by following these five:

1. Yoshua Bengio – Understand AI safety.
2. Stuart Russell – Understand human-compatible AI.
3. Timnit Gebru – Understand ethical AI.
4. Joy Buolamwini – Understand AI bias and fairness.
5. Shannon Vallor – Understand the philosophy of AI ethics.

One important piece of advice: Follow people who disagree with each other. AI safety is not a single school of thought. You'll become a much stronger critical thinker by understanding competing perspectives.

One question to help me guide you further

What is your main goal in learning AI Safety and Ethics?

To learn about AI safety and ethics, it helps to look at a mix of researchers, public intellectuals, advocacy groups, and academic institutions. Because the field spans everything from immediate societal harms (bias, discrimination, privacy, misinformation) to long-term existential risks (AI alignment, superintelligence control), the best approach is to follow a diverse group of voices.

Key people, organizations, and resources to follow include:

1. Prominent Thinkers & Researchers

• Timnit Gebru: Founder of the Distributed AI Research Institute (DAIR). A leading voice on algorithmic bias, ethical data collection, and the social harms of commercial AI development.

• Stuart Russell: UC Berkeley computer science professor and author of Human Compatible AI. A pioneer in technical AI safety and value alignment (ensuring AI goals remain aligned with human values).

• Yoshua Bengio: Turing Award winner and head of Mila (Quebec AI Institute). He has become a prominent advocate for government regulation, whistleblower protections, and pausing reckless development of advanced systems.

• Joy Buolamwini: Founder of the Algorithmic Justice League and author of Unmasking AI. Essential to follow for insights on racial justice, accountability, and the civil rights implications of automated systems.

• Max Tegmark: Physicist at MIT and president of the Future of Life Institute. Focuses heavily on the long-term existential risks of artificial general intelligence (AGI) and global governance frameworks.

• Merve Hickok: Founder of AIethicist.org and President of the Center for AI & Digital Policy. Great for practical governance, regulatory updates, and corporate accountability.

• Emily M. Bender: University of Washington linguistics professor who frequently critiques "hype," large language model limitations, and the ethical shortcuts taken by major tech companies.

2. Key Organizations & Research Institutes

• DAIR (Distributed AI Research Institute): Focuses on independent, community-rooted AI research that counters Big Tech narratives and prioritizes marginalized groups.

• Center for Human-Compatible AI (CHAI): Based at UC Berkeley, dedicated to re-orienting AI research toward provably beneficial systems. Academic Organizations(24) - The Directory of AI Security & Safety

• Stanford HAI (Human-Centered AI): Publishes the annual AI Index and bridges technical progress with public policy, ethics, and human impact.

• The Alan Turing Institute (UK): The United Kingdom’s national institute for data science and AI, which heavily produces guidelines on safe, ethical, and trustworthy AI use. Academic Organizations(24) - The Directory of AI Security & Safety

• Future of Life Institute (FLI): Known for organizing open letters calling for cautious AI development and researching catastrophic risk mitigation.

3. Newsletters, Podcasts, and Aggregators

If you want regular curation delivered to your inbox or feed, look into these:

• The Gradient: A publication offering accessible breakdowns of AI research, ethics, and safety from practitioners.

• AI Snake Oil (Princeton): A Substack by Arvind Narayanan and Sayash Kapoor that cuts through AI hype and rigorously evaluates claims about AI capabilities and societal impacts.

• The Robot Brains Podcast / PAIR (People + AI Research at Google): Excellent entry points for understanding how AI interacts with human design and cognitive psychology.

Are you looking to focus more on immediate ethical concerns (like bias, copyright, and employment) or long-term technical safety (like AI alignment and existential risk)? Let me know and I can tailor more specific reading recommendations!

Learning about AI safety and ethics requires looking at a mix of technical alignment (preventing catastrophic loss of control) and societal ethics (bias, governance, justice, and accountability).

The field is generally split between technical safety researchers and sociotechnical ethics/governance experts. Top figures and organizations to follow include:

Sociotechnical Ethics, Fairness, & Governance

** Dr. Rumman Chowdhury:** A leading expert in responsible AI, former head of ML Ethics, Transparency, and Accountability at Twitter/X, founder of Humane Intelligence, and a U.S. Science Envoy for AI. She focuses heavily on audits, red teaming, and practical governance. Dr. Rumman Chowdhury - How to FIX AI Before It's Too Late?

** Dr. Timnit Gebru:** Computer scientist, founder of Black in AI, and founder/director of the Distributed AI Research Institute (DAIR). She is renowned for her work exposing bias in facial recognition and critiquing Big Tech's power structures and ethics washing. Timnit Gebru: Ethical AI Requires Institutional and Structural Change | Stanford HAI

** Dr. Margaret Mitchell:** An ethical AI researcher and former co-lead of Google’s Ethical AI team, now working as a chief ethics scientist. She frequently writes and speaks on transparency, accountability, and the real-world harms of generative models.

** Dr. Deborah Raji:** A researcher at UC Berkeley focusing on algorithmic auditing, evaluation, and accountability, known for her ground-breaking work on bias in commercial facial analysis systems.

Technical Alignment & Existential Safety

** Paul Christiano:** A pioneer in AI alignment (co-inventor of RLHF) and founder of the Alignment Research Center (ARC). He focuses deeply on what scalable oversight and failure modes look like as models get smarter. Top 30 AI Safety Researchers. The following is a list of the top 30… | by Kevin O'Shaughnessy | Aug, 2026 | MediumBlunt Warnings from an AI Researcher Who Just Quit

** Jan Leike:** A prominent alignment researcher who formerly co-led superalignment at OpenAI and moved to Anthropic, focusing on how to steer and align advanced AI systems. Top 30 AI Safety Researchers. The following is a list of the top 30… | by Kevin O'Shaughnessy | Aug, 2026 | Medium

** Stuart Russell:** Computer scientist at UC Berkeley and author of Human Compatible AI. He is the primary voice behind designing AI systems that are provably aligned with human values rather than optimizing blind proxy goals. Top 30 AI Safety Researchers. The following is a list of the top 30… | by Kevin O'Shaughnessy | Aug, 2026 | Medium

** Rohin Shah:** A researcher at Google DeepMind who leads AGI safety alignment and previously curated the widely read Alignment Newsletter, which maps out technical progress in the field. Top 30 AI Safety Researchers. The following is a list of the top 30… | by Kevin O'Shaughnessy | Aug, 2026 | Medium

Organizations & Newsletters to Follow

• Center for AI Safety (CAIS): Directed by Dan Hendrycks, known for publishing foundational papers on AI risks, benchmarks, and policy statements signed by leading scientists.

• DAIR (Distributed AI Research Institute): Focuses on independent, community-rooted research on the ethical impacts of AI.

• Alignment Forum: The central online hub where technical researchers actively debate alignment math, mechanistic interpretability, and agent architectures. Timnit Gebru: Ethical AI Requires Institutional and Structural Change | Stanford HAITop 30 AI Safety Researchers. The following is a list of the top 30… | by Kevin O'Shaughnessy | Aug, 2026 | Medium

Would you prefer to focus more on the technical/math side of alignment (like model steerability and interpretability) or the policy/societal side (like bias, labor, and regulation)? I can point you toward specific reading lists or papers.

Follow a mix of voices across technical alignment and social impact: Stuart Russell and Dan Hendrycks for technical safety; Timnit Gebru, Emily Bender, and Margaret Mitchell for ethics and societal risks; and Shannon Vallor for philosophy and responsible technology. Robert Miles is a good starting point for accessible technical explainers.(citegist.com)(aisafety.com)(aiweekly.co)

To learn about AI safety and ethics, follow a mix of technical researchers, policy leaders, and civil society voices. This ensures you get perspectives on both the technical risks of advanced AI and the broader societal implications of its deployment.

🌐 Key Individuals in AI Safety & Ethics

🏛️ Organizations to Watch

• Machine Intelligence Research Institute (MIRI) – Foundational AI safety research.Resource Directory - Organizations, Papers, Books & Courses on AGI Risk | AGI Risk

• Center for AI Safety (CAIS) – Advocacy and technical research on catastrophic risks.Resource Directory - Organizations, Papers, Books & Courses on AGI Risk | AGI Risk

• Centre for the Governance of AI (GovAI) – Policy-focused research on AI governance.Resource Directory - Organizations, Papers, Books & Courses on AGI Risk | AGI Risk

• Future of Life Institute – Global advocacy on existential risks.Resource Directory - Organizations, Papers, Books & Courses on AGI Risk | AGI Risk

• Ada Lovelace Institute – Public interest research on AI ethics.Responsible ai Organizations in AI Safety (111) | AI Safety Directory

• AI Now Institute – Examines AI’s societal implications, including bias and labor.Responsible ai Organizations in AI Safety (111) | AI Safety Directory

⚠️ Risks & Trade-offs

• Fragmentation of perspectives: Technical safety researchers (e.g., alignment labs) often disagree with policy experts on timelines and risks. Following across camps helps avoid echo chambers.People to follow in AI safety · AI Safety Guide

• Hype vs. substance: Some voices focus on speculative existential risks, while others emphasize immediate harms like bias, surveillance, and labor impacts. Balance both.

• Global diversity: Much discourse is US/Europe-centric. Including voices from Africa, Asia, and Latin America (e.g., UNESCO’s AI Ethics Experts Without Borders) broadens perspective.AI Ethics Experts Without Borders - List

✅ Actionable Guide

1. Follow technical researchers (Russell, Christiano, Leike, Hendrycks) for alignment and interpretability.

2. Track governance experts (Toner, Véliz, Renieris, EU AI Act drafters like Lucilla Sioli) for regulation updates.88 AI Governance and AI Regulation experts to follow (EU AI Act, US, China, UN) - AI Governance Library, Articles, Ressources

3. Engage with civil society groups (Ada Lovelace Institute, AI Now, Algorithmic Justice League) for ethical and social perspectives.Responsible ai Organizations in AI Safety (111) | AI Safety Directory

4. Balance feeds: Include both existential risk advocates (Tegmark, Future of Life Institute) and near-term harm researchers (AI Forensics, AI Incident Database).Responsible ai Organizations in AI Safety (111) | AI Safety Directory

Would you like me to curate a “starter list” of Twitter/X and LinkedIn profiles so you can follow these experts directly? That way, you’ll have a ready-made feed of diverse voices in AI safety and ethics.