What is the average AI Safety and Evaluation salary?

$277K a year on average and $265K at the median, across 134 open AI Safety and Evaluation postings at companies that state pay (November 2026). The median is the better guide: a few very high offers pull the average up.

What is the typical range?

The middle half of the postings offer $212K to $326K a year. A quarter offer less and a quarter offer more.

Which companies advertise the highest AI Safety and Evaluation pay?

Anthropic ($362K), Harvey ($340K), OpenAI ($332K), by the median of their open postings that state pay (three postings or more).

Is an ai safety and evaluation a high-paying job?

The median AI Safety and Evaluation posting advertises $265K a year, more than the median of all 55,899 job postings with pay on Jobs Every Hour ($174K). That is 52% above the middle of the market.

How much can you earn as an ai safety and evaluation?

A tenth of the postings offer $402K a year or more; the highest open AI Safety and Evaluation role advertises up to $850K. Pay rises with level: see the table by level.

What is the AI Safety and Evaluation salary per hour and per month?

The median of $265K a year is about $127 an hour (2,080 working hours) and $22,083 a month before tax. The middle half of postings: $102 to $157 an hour.

Which skills pay more for AI Safety and Evaluation roles?

Postings that ask for AI agents (+$48K, 13 postings), Machine Learning (+$31K, 29 postings), Generative AI (+$5K, 13 postings) advertise more than the $265K median of all AI Safety and Evaluation postings with pay.

Where do these figures come from?

From the pay ranges employers print in their own job postings, read from their careers pages every hour. The midpoint of each range is counted. These are advertised ranges for open roles, mostly in the United States and in US dollars, not what people are actually paid.

AI Safety and Evaluation Salary: $277K Average, $265K Median

AI Safety and Evaluation salary by level

Lead roles advertise a median of $280K, 33% more than senior roles ($210K). The level comes from the job title; titles without a level word are counted separately.

Level	Median salary	Middle half	Roles
Senior	$210K	$202K–$224K	18
Lead and above	$280K	$256K–$380K	25
No level in the title	$266K	$216K–$340K	91

AI Safety and Evaluation salary by location

Name	Median salary	Roles counted
🇺🇸New York, NY	$308K	29
🇺🇸San Francisco Bay Area	$276K	96
🇺🇸California	$275K	101
🇺🇸United States	$265K	127
🇺🇸Boston, MA	$245K	15

Skills that pay more for AI Safety and Evaluation roles

Skill	Median with it	Above the median	Postings
AI agents	$312K	+$48K	13
Machine Learning	$296K	+$31K	29
Generative AI	$270K	+$5K	13

Top-paying companies

Name	Median salary	Roles counted
Anthropic	$362K	31
Harvey	$340K	3
OpenAI	$332K	12
Mercor	$279K	4
Perplexity	$275K	3
Lila Sciences	$264K	8
Waymo	$238K	4
Anduril Industries	$213K	5
LangChain	$208K	10

Highest-paying open AI Safety and Evaluation roles

All AI safety and evaluation jobs

Research Engineer, Post-Training Model EvaluationsAnthropic · 🇺🇸San Francisco, CA | Seattle, WA$500K - $850KPythonLLMLLM evaluation6mo
Research Scientist, InterpretabilityAnthropic · 🇺🇸San Francisco, CA$350K - $850KPythonLLM12mo
Research Scientist / Research Engineer, Safety + Alignment, SearchOpenAI · 🇺🇸San Francisco$380K - $850K3w
Member of Technical Staff, EvalsMagic.dev · 🇺🇸San Francisco$200K - $550KRemoteMachine LearningLLM evaluation4d
Researcher, AlignmentOpenAI · 🇺🇸San Francisco$295K - $500KPythonTypeScriptMachine Learning27mo
Tech Lead, Applied AI - EvalsMercor · 🇺🇸San Francisco$260K - $500KLead and aboveGoPythonTypeScript5d
Member of Technical Staff - Mechanistic InterpretabilityVmax · 🇺🇸SF$300K - $500KPythonMachine LearningLLM5mo
Research Engineer / Scientist, Data & EvalsHarvey · 🇺🇸San Francisco$230K - $450KHybridPythonLLMAI agents13d
Research Engineer / Scientist, Data & EvalsHarvey · 🇺🇸New York$230K - $450KHybridPythonLLMAI agents13d
Member of Technical Staff, AlignmentAbundant · 🇺🇸San Francisco$250K - $450KLLM2mo
Member of Technical Staff, Enterprise Evals PlatformMercor · 🇺🇸San Francisco$220K - $425KLLMLLM evaluationStatistics3mo
Member of Technical Staff (Software Engineer, Enterprise Agent Evals & Efficiency)Perplexity · 🇺🇸San Francisco$220K - $405KRustGoPython3w
Research Engineer, AI Safety & AlignmentCharacter.AI · 🇺🇸Redwood City, CA$225K - $400KHybridDockerMachine LearningLLM13mo
Member of Technical Staff, Model EvaluationMirendil · 🇺🇸San Francisco$300K - $400KLLM evaluationObservability4mo
Staff Product Manager - AI EvalsTwenty · 🇺🇸Arlington, VA$161K - $389KLead and aboveAISecurity clearanceAI agents3w
Senior / Principal Chemist, AI SafetyLila Sciences · 🇺🇸🇬🇧Cambridge, MA USA; London, UK; San Francisco, CA USA$268K - $384KLead and aboveAI5w
Senior / Principal Computational Biologist, AI SafetyLila Sciences · 🇺🇸🇬🇧Cambridge, MA USA; London, UK; San Francisco, CA USA$268K - $384KLead and aboveLLMAI3w
Tech Lead, Self Driving Eval InfrastructureWaymo · 🇺🇸Mountain View, CA, USA$281K - $356KLead and abovePythonJavaC++11mo
Member of Technical Staff (Answer Quality & Evals)Perplexity · 🇺🇸San Francisco$200K - $350KHybridPythonSQLMachine Learning7mo
Staff Technical Product Manager, EvalsHercules · 🇺🇸San Francisco$200K - $350KLead and aboveTypeScriptReactPostgreSQL3w

Questions about AI Safety and Evaluation pay

What is the average AI Safety and Evaluation salary?: $277K a year on average and $265K at the median, across 134 open AI Safety and Evaluation postings at companies that state pay (November 2026). The median is the better guide: a few very high offers pull the average up.
What is the typical range?: The middle half of the postings offer $212K to $326K a year. A quarter offer less and a quarter offer more.
Which companies advertise the highest AI Safety and Evaluation pay?: Anthropic ($362K), Harvey ($340K), OpenAI ($332K), by the median of their open postings that state pay (three postings or more).
Is an ai safety and evaluation a high-paying job?: The median AI Safety and Evaluation posting advertises $265K a year, more than the median of all 55,899 job postings with pay on Jobs Every Hour ($174K). That is 52% above the middle of the market.
How much can you earn as an ai safety and evaluation?: A tenth of the postings offer $402K a year or more; the highest open AI Safety and Evaluation role advertises up to $850K. Pay rises with level: see the table by level.
What is the AI Safety and Evaluation salary per hour and per month?: The median of $265K a year is about $127 an hour (2,080 working hours) and $22,083 a month before tax. The middle half of postings: $102 to $157 an hour.
Which skills pay more for AI Safety and Evaluation roles?: Postings that ask for AI agents (+$48K, 13 postings), Machine Learning (+$31K, 29 postings), Generative AI (+$5K, 13 postings) advertise more than the $265K median of all AI Safety and Evaluation postings with pay.
Where do these figures come from?: From the pay ranges employers print in their own job postings, read from their careers pages every hour. The midpoint of each range is counted. These are advertised ranges for open roles, mostly in the United States and in US dollars, not what people are actually paid.

AI Safety and Evaluation salary: $265K median