OpenAI engineers can earn up to $800K per year — among the highest rates in the industry

OpenAI Careers
(Image credit: OpenAI)

OpenAI engineering staff are among the best paid in the industry, according to an investigation by Bloomberg. Data suggests that a talented recruit at the ChatGPT firm can expect salary plus bonuses and stock awards to be worth about $800,000 per year. We looked at earlier Artificial Intelligence (AI) engineer salary data, and the recent report seems to confirm that OpenAI is the biggest payer in AI, with new engineers and researchers netting up to twice as much as they could from working at Nvidia, for example.

AI is more than just an IT industry buzzword in 2023, and the year might be remembered as one where the foundations of generative AI as we know it were laid. Companies with their fingers in AI pies have been great earners for investors this year, and this good fortune appears to trickle down to their employees – with pay rates for those employed for their AI expertise rocketing.

Bloomberg’s report (citing Levels.fyi data) says that AI engineers can expect up to 12.5% better salaries than non-AI counterparts. If you check out the engineering job vacancies at OpenAI, you might see a broad range of salaries ranging from $200K to $450K. However, adding in bonuses and stock awards, lucky employees will enjoy an annual salary closer to $800K in total.

The reason for the lucrative salaries (and bonuses), appears to be the old law of supply and demand in action. According to the source report, it is extremely difficult to get experienced and appropriately skilled people – and to retain them. Moreover, experience is said to be especially important here as “there’s a big difference between understanding AI models on a theoretical level and having the skills and experience to actually apply them,” states Bloomberg.

The AI talent market doesn’t look set to enjoy any pressure relief in the medium term. We still appear to be riding high on the AI wave, with 2024 already unofficially being called the year of the AI PC. Thus, AI skills are going to be in even stronger demand to develop new apps and experiences to propel all the new AI hardware and software like Windows 12.

OpenAI looks to raise new funding at $100B valuation

In other OpenAI news, there are reports that the firm is seeking to raise new funds at a market valuation of $100B or more. The source says details of the fundraising are still not set in stone, so terms, timings, and valuations could change. There is also talk of a new chip venture between OpenAI and Abu Dhabi-based G42.

Mark Tyson
Freelance News Writer

Mark Tyson is a Freelance News Writer at Tom's Hardware US. He enjoys covering the full breadth of PC tech; from business and semiconductor design to products approaching the edge of reason.

  • USAFRet
    "a talented recruit can expect"

    This is like being the first or second round draft pick in the NFL/NBA.

    Not your typical person.


    One of the little chuckleheads at my office is going to see this, and assume it means him personally.
    "No dude, you are not that."
    Reply
  • jeremyj_83
    Better question is where can someone get the training to get into this field of IT.
    Reply
  • bit_user
    USAFRet said:
    "a talented recruit can expect"

    This is like being the first or second round draft pick in the NFL/NBA.

    Not your typical person.
    Yeah, they surely want folks with advanced degrees from universities with the top AI programs or their leading industry competitors. If you just dabble with AI at home and take a free online course, don't even expect to hear back from them after submitting your resume'.

    It reminds me of the kinds of Wall St. jobs I heard about, 10-15 years ago. If you had a double major in CS and finance, from a top university, you could land a job at a hedge fund and make $350k. The average developer need not apply.

    jeremyj_83 said:
    Better question is where can someone get the training to get into this field of IT.
    It's not hard to find online classes. They'll give you a basic grounding in the theoretical underpinnings, plus teach you the main points about how to apply AI to tackle different sorts of problems. That might get your foot in the door, at most places offering more normal salaries - especially if you have some adjacent work experience to draw from.

    I took Andrew Ng's original Machine Learning course, at Coursera, not long after it launched. It was derived directly from his Stanford course, but they watered down the math and (I think) did a bit more hand-holding through the exercises. I'm not sure which of their current courses it is, but perhaps they split it into multiple ones:
    https://www.coursera.org/search?query=AI&partners=Stanford+University
    For me, the time commitment of those courses was the biggest problem. I don't know if it's the same as before, but it's no joke. You'd think online learning means you get to go at your own pace, but they kept it to a normal university class calendar + 1 week grace period.
    Reply
  • slightnitpick

    can earn up to
    People, including me, use the word "earn" for salaries all the time. But I do wish editors and journalists would normalize to "be paid" or "compensated" or the like. As a word, "earn" brings with it a moral judgement of worth or entitlement to both no more and no less than the sum so earned. "Get" saves a letter for a title.

    And we all know in real life, much less capitalism, that this is not the case. Compensation comes down to negotiation, with employees usually paid less than they earn, so that the capital (or in the case of a non-profit, the non-profit mission) has a return on investment. And sometimes employees are paid more than they earn, especially in the case of job-hoppers, but also in the case of start-ups and other scenarios where the pay is a gamble or there just aren't enough people to do a job at the value-added wage.

    I know there are people who believe that this is just, but I think even they would typically agree that, at least in well-functioning organizations, an employee has to "earn" more value for the organization than they are paid. Otherwise the organization will ultimately fail.
    Reply
  • PEnns
    jeremyj_83 said:
    Better question is where can someone get the training to get into this field of IT.

    And the best part? They always want you to have X years of experience in such a new field!!
    Reply
  • Zemeckis5
    This is incorrect information. Engineers at OpenAI can earn well over $800k not “up to”. Industry hires start at around $1M and it only goes up from there. $800k might be the upper end for someone fresh out of a PhD with no industry experience
    Reply
  • USAFRet
    Zemeckis5 said:
    This is incorrect information. Engineers at OpenAI can earn well over $800k not “up to”. Industry hires start at around $1M and it only goes up from there. $800k might be the upper end for someone fresh out of a PhD with no industry experience
    6 months ago, starting salaries may have been different.
    6 months from now, different again.
    Reply
  • bit_user
    Zemeckis5 said:
    This is incorrect information. Engineers at OpenAI can earn well over $800k not “up to”. Industry hires start at around $1M and it only goes up from there. $800k might be the upper end for someone fresh out of a PhD with no industry experience
    Source?
    Reply