Meta tags:
description=Apollo Research is focused on reducing risks from scheming frontier AI. We conduct fundamental research into the science of scheming and run pre-deployment evaluations of frontier AI systems. Our governance team provides global policymakers with expert technical guidance, and our product team builds AGI safety tools to monitor and secure frontier AI agents.;

Headings (most frequently used words):

ai, scheming, of, and, the, for, research, to, risks, from, our, science, apollo, monitoring, systems, that, reducing, frontier, mission, watcher, collaborations, other, inquiries, please, get, in, touch, governance, policy, products, loss, control, threat, map, development, we, need, update, may, 2026, scalable, agenda, stress, testing, deliberative, alignment, anti, training, as, capabilities, increase, some, greatest, will, come, advanced, covertly, pursue, misaligned, objectives, goal, is, understand, evaluate, emergence, prevent, possible, harms, might, cause, build, tools, make, deployment, powerful, safer, agent, tool, by,

Text of the page (most frequently used words):
and (17), #science (10), more (10), scheming (10), read (9), research (8), the (8), apollo (7), our (7), #products (6), governance (6), policy (5), for (5), frontier (5), that (5), 2026 (5), systems (5), agent (4), from (4), watcher (4), contact (3), about (3), get (3), touch (3), with (3), misaligned (3), may (3), how (3), monitoring (3), risks (3), detect (3), now (2), cookie (2), privacy (2), press (2), blog (2), careers (2), other (2), organizations (2), scale (2), please (2), out (2), models (2), covertly (2), stress (2), tests (2), environments (2), training (2), are (2), evaluation (2), alignment (2), goal (2), this (2), agenda (2), into (2), safety (2), best (2), coding (2), agents (2), advanced (2), building (2), prevent (2), threat (2), map (2), loss (2), control (2), tools (2), deployment (2), designed, transparency, register, 837878591690, ein, 4310599, usa, pbc, interested, partnering, you, have, lot, logs, would, like, analyze, them, automatically, reach, collaborations, inquiries, partners, september, 2025, partnered, openai, assess, language, early, signs, pursuing, goals, controlled, non, typical, studied, method, can, significantly, reduce, but, not, eliminate, these, behaviors, results, complicated, increasing, ability, recognize, their, testing, deliberative, anti, translate, compute, case, figure, spend, order, 100m, produce, extremely, good, monitors, scalable, has, office, san, francisco, hiring, across, many, roles, including, update, post, january, face, strong, incentives, scheme, predict, risk, need, june, created, traces, pathways, use, automation, towards, outcomes, development, latest, learn, emergent, manipulation, data, exfiltration, insecure, code, execution, catches, dangerous, behavior, before, becomes, incident, works, tailscale, aperture, tool, agi, monitor, secure, first, product, automated, oversight, layer, detects, failure, modes, real, time, support, governments, international, developing, technical, regimes, enable, effective, regulation, establish, standards, practices, conduct, fundamental, emerges, mitigate, run, pre, evaluations, strategic, deception, awareness, behaviour, understand, evaluate, emergence, possible, harms, might, cause, build, make, powerful, safer, capabilities, increase, some, greatest, will, come, pursue, objectives, mission, reducing, team,

Text of the page (random words):
apollo research about team blog press contact us science governance products careers reducing risks from scheming frontier ai science governance products about us read more our mission as ai capabilities increase some of the greatest risks will come from scheming ai advanced systems that covertly pursue misaligned objectives our goal is to understand and evaluate the emergence of scheming to prevent the possible harms that scheming ai might cause and build tools to make the deployment of powerful ai systems safer science we conduct fundamental research into the science of scheming how it emerges and how to detect and mitigate it we run pre deployment evaluations of frontier ai systems to detect strategic deception evaluation awareness and misaligned behaviour read more governance policy we support governments and international organizations by developing technical ai governance regimes we enable effective regulation of frontier ai systems and establish standards and best practices read more products we are building agi safety tools to monitor and secure frontier ai agents our first product watcher is an automated oversight layer that detects failure modes in real time read more our products watcher ai agent monitoring tool by apollo research watcher catches dangerous coding agent behavior before it becomes an incident watcher works with tailscale aperture to detect insecure code execution data exfiltration agent manipulation emergent risks learn more latest from apollo governance a loss of control threat map for ai research and development we created a threat map that traces pathways from ai use in ai r d automation towards loss of control outcomes june 17 2026 read more science we need a science of scheming advanced ai systems face strong incentives to scheme apollo research is building a science of scheming to predict and prevent this risk january 19 2026 read more post apollo update may 2026 apollo research now has an office in san francisco and is hiring across many roles including science of scheming and monitoring may 13 2026 read more products a scalable monitoring research agenda the goal of this agenda is to translate compute into safety at scale in the best case we figure out how to spend on the order of 10 100m to produce extremely good monitors for coding agents may 8 2026 read more science stress testing deliberative alignment for anti scheming training we partnered with openai to assess frontier language models for early signs of scheming covertly pursuing misaligned goals in controlled stress tests non typical environments and studied a training method that can significantly reduce but not eliminate these behaviors our results are complicated by models increasing ability to recognize our evaluation environments as tests of their alignment september 17 2025 read more our partners contact us get in touch for collaborations and other inquiries please get in touch we re interested in partnering with other organizations if you have a lot of ai agent logs and would like to analyze them automatically at scale please reach out get in touch apollo research pbc ein 93 4310599 de usa eu transparency register 837878591690 80 privacy policy cookie policy science governance products careers about blog press contact us privacy policy cookie policy designed by and now