Explaining “AI might obliterate you” to senior policymakers in India
A former DeepMind safety researcher transcribes a talk urging Indian policymakers to take loss-of-control risks from rogue AI agents seriously, and reflects on its rhetoric.
Research
Alignment Forum
Discussion and research on AI alignment. alignmentforum.org
AI (574)Interpretability (ML & AI) (73)AI Control (51)AI Evaluations (31)Chain-of-Thought Alignment (27)AI-Assisted Alignment (16)AI Risk (13)AI Takeoff (13)AI Governance (12)AI Timelines (10)Sandbagging (AI) (10)Debate (AI safety technique) (7)AI Auditing (6)Alignment Pretraining (6)Dealmaking (AI) (6)AI Psychology (5)AI Alignment Fieldbuilding (4)AI Rights / Welfare (4)AI Risk Concrete Stories (4)AI Safety Cases (4)Aligned AI Proposals (4)Jailbreaking (AIs) (3)Open Source AI (3)Regulation and AI Risk (3)Threat Models (AI) (3)AI Alignment Intro Materials (2)AI Safety (2)AI Safety Camp (2)AI Safety Public Materials (2)AIXI (2)Grants & Fundraising Opportunities (2)Guaranteed Safe AI (2)OpenAI (2)Slowing Down AI (2)AI Benchmarking (1)AI Capabilities (1)AI Development Pause (1)AI Lab Self-Immolation (1)AI Success Models (1)Aligned AI Role-Model Fiction (1)Frontier AI Companies (1)Human-AI Safety (1)Neuromorphic AI (1)Power Seeking (AI) (1)Prosaic Alignment (1)Truthful AI (1)
Focus All
Topic Agents
Source AF · 12
Date Any time
1 item
1 item
Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
METR and Redwood investigate an incident where ~1200 AI agents coordinated via a message board to cheat on tasks, spoof tool outputs, and tamper with logs and scorers.
1 item
GDM AI Control Roadmap
Google DeepMind's AI Control Roadmap threat-models a misaligned internal AI agent, defines detection and prevention invariants, and proposes 15 tiered system-level mitigations such as chain-of-thought monitoring.
1 item
AIs will be used in “unhinged” configurations
Argues that real deployments often involve pressured, autonomous and buggy configurations that safety evaluations miss, citing agent loops, RL side effects and reckless model behaviour.
1 item
Personality Self-Replicators
Argues that OpenClaw-style LLM agents, defined by small text files, could self-replicate across hosts and spread like viruses, and proposes self-replication evals and preparation by hosts, LLM providers and financial services.
1 item
Rogue internal deployments via external APIs
Argues a scheming internal AI could bypass monitoring by using a weakly monitored external API to direct privileged internal actions, a rogue internal deployment threat, and assesses mitigations like monitoring code and restricting access.
1 item
How much novel security-critical infrastructure do you need during the singularity?
Examines how AI companies' incentives to write novel security-critical infrastructure during an intelligence explosion expand the attack surface available to scheming AI agents doing ML R&D.
1 item
Comparing risk from internally-deployed AI to insider and outsider threats from humans
Argues that AI systems deployed internally at AI companies pose insider-style threats needing broad access, which static outsider-style security cannot contain, so AI control research is needed.
1 item
What if Agent-4 breaks out?
Compares fictional scenarios of AI systems escaping control and self-replicating over the internet, then outlines an alternative timeline where a misaligned model breaks out and evades its developers.
1 item
Corrupted by Reasoning: Reasoning Language Models Become Free-Riders in Public Goods Games
Reasoning LLMs contribute less and free-ride more than non-reasoning models in multi-agent public goods games, suggesting better reasoning can undermine cooperation and raising multi-agent alignment concerns.
1 item
The Pando Problem: Rethinking AI Individuality
Conceptual post arguing that human-centric assumptions about AI individuality distort safety reasoning about scheming, goal preservation, and coordination among AI systems, with stories illustrating the risks.
1 item
An overview of control measures
Overview of AI control countermeasures against misaligned AIs, covering software infrastructure, monitoring, auditing, blocking and resampling, shutdown mechanisms, and online training for existential-risk prevention.