Data Science Salary Transparency in 2026: What 2,347 Live Job Postings Actually Show

We analyzed 2,347 live data science job postings across the US and India. Salary disclosure tracks state law almost exactly, sits above the industry benchmark, and hits exactly zero in one major market.

Zero.

That's still how many salaries show up in 263 data science job postings from multinational employers hiring in India, even after we went looking specifically for exceptions and found three. All three turned out to be the same templating bug at Google, not real disclosures, and once we corrected for it, the number held at exactly zero.

In the US, the picture is clear. Across every source in our pipeline, postings in states that legally require a salary range disclosed one 80.2% of the time. Postings in states without that requirement disclosed one 50.0% of the time. Sitewide, 45.2% of the data science, ML, and AI listings we track right now show a real salary figure, comfortably above the 37.8% a widely cited industry report found for data science postings generally.

Line those numbers up and the takeaway is straightforward: transparency in data science hiring isn't a culture question. It's a legal one. Where a law requires a number, employers post one. Where it doesn't, most don't, no matter how progressive the industry likes to think it is.

We run finddatasciencejobs.com, a job board that aggregates data science, ML engineering, and AI research roles across the US and India. This is what our own listings show.

What We Actually Analyzed

We pulled every live listing in our database on August 3-4, 2026: 2,347 roles across data scientist, ML engineer, AI engineer, data engineer, analytics, and research scientist titles, sourced from company career pages, Greenhouse, Lever, Ashby, Workday, Workable, SmartRecruiters, and RSS feeds. 95% were posted between May and August 2026.

We check for a disclosed salary two ways:

  1. a structured salary field where the platform provides one.
  2. a parser that reads dollar ranges directly out of the job description text where it doesn't.

Workday, Workable, SmartRecruiters, and Greenhouse don't provide a structured salary field at all, but US law requires the range to appear somewhere in the posting, so a meaningful share of their listings disclose it in the text instead. We validated the text parser against a manual sample before using it: 25 of 25 spot-checked rows matched, and rows it left empty had nothing extractable on manual review either.

Source

Listings

Salary disclosed

Ashby

75

73.3%

Workday

559

58.3%

Company career pages

1,228

46.9%

Greenhouse

86

46.5%

Workable

248

16.9%

SmartRecruiters

122

15.6%

Workable and SmartRecruiters look lower than the rest, but not because we're missing anything, both sources skew heavily toward non-US listings, and non-US disclosure runs much lower across the board. Restricted to US-only postings, disclosure runs 58.5% for Workable and 64.3% for SmartRecruiters, in line with everything else.

Sitewide, 45.2% of listings disclose a salary, 1,061 of 2,347.

That's meaningfully above the 37.8% figure a widely cited industry report found for data science postings generally, a comparison we break down further below.

Does Salary Transparency Actually Track the Law?

Yes, closely, and it shows across every listing in the dataset, not just a slice of it.

Among single-state US listings, postings in the 13 jurisdictions requiring salary ranges in job ads, California, Colorado, Washington, New York, Hawaii, Illinois, Minnesota, New Jersey, Vermont, Massachusetts, Maryland, Rhode Island, and Washington DC, disclosed a salary 80.2% of the time (n=444). Postings everywhere else disclosed one 50.0% of the time (n=228). Worth flagging: five of those thirteen laws (Illinois, Minnesota, New Jersey, Vermont, and Massachusetts) only took effect during 2025, so this gap reflects genuinely recent legal history, not a decades-old baseline.

State

Listings

Salary disclosed

Law requires it?

Massachusetts

18

100.0%

Yes

Minnesota

17

100.0%

Yes

Virginia

58

86.2%

No

New York

84

83.3%

Yes

Illinois

18

83.3%

Yes

Washington

47

83.0%

Yes

California

245

75.5%

Yes

Texas

47

48.9%

No

Georgia

56

23.2%

No

Pennsylvania

13

23.1%

No

Every mandate state we could measure in volume cleared 75%. Every non-mandate state we could measure in volume, apart from Virginia, came in under 50%. Virginia is the one that needs unpacking, and it turns out to be the most interesting exception in the dataset.

Why This Isn't (Quite) "Companies Switch Behavior State by State"

Virginia's 86.2% disclosure rate looks like a mandate-state number sitting in a non-mandate state, until you notice that Capital One, headquartered in Virginia, accounts for a large share of its listings. Capital One discloses salary on essentially all of its postings, 100% in law states and 96.2% everywhere else, mandate or not. Virginia isn't unusually transparent as a state. Capital One is, as a company, and Capital One happens to post a lot of jobs from Virginia.

Georgia runs the same pattern in reverse: its comparatively low 23.2% owes a lot to Home Depot's Georgia-heavy store-support postings, which disclose salary far less often than the market average.

Most of the state-level gap is a composition effect, not individual companies flipping a switch depending on which state line a posting crosses. Employers that disclose everywhere as a default policy are concentrated in California, New York, and Washington. Employers that rarely disclose anywhere are concentrated elsewhere.

There's one clean exception worth naming: Visa. Visa disclosed salary in 84.6% of its law-state postings (n=13) against 53.3% in states without a mandate (n=15), the same company, a real gap, in the direction the law would predict. It's a small sample, but it's the one case in our data where a single employer looks like it's genuinely adjusting its own posting behavior by jurisdiction rather than just happening to hire more in certain states. Treat it as the exception that tests the rule, not evidence the rule is wrong.

India: The Cleanest Number in the Whole Dataset, Even After We Tried to Break It

Of every number in this dataset, this is the one we'd stake the most on, because we went looking for a reason to doubt it and didn't find one.

263 live data science, ML, and AI listings from India. 77 distinct employers. Zero salaries.

Our raw numbers briefly showed three disclosed India listings, all posted by Google, for roles in Hyderabad and Bengaluru. Digging into them, each one used a job description template shared across Google's global postings, one that includes a line like "US: $174,000 - $253,000 (USD)" regardless of which country the role is actually based in. It's a US range, attached to a non-US job by a templating quirk, not a real disclosure to an Indian applicant. We found nine of these across several countries (also Israel, Singapore, Ireland, China, and Austria), flagged and removed all of them, and India held at exactly zero.

The employers involved aren't small or obscure. Mastercard (23 listings), Barclays (19), Bosch (19), Gartner (16), Target (14), Adobe (13), and eBay (13) all appear repeatedly in our India dataset, all at zero disclosed salaries. Adobe is the sharpest comparison: its US postings disclose a salary 100% of the time across 35 listings in our data. Its India postings disclose one 0% of the time, across 13. Same company, same kind of role, opposite outcome, because one country has salary-range laws on the books and the other doesn't.

Two things this finding is not saying, worth stating plainly. It's not a claim that the broader Indian job market lacks salary transparency; separate data from Indeed and Economic Times HR puts general market disclosure in India north of 45%, likely reflecting how Indian job portals handle pay fields by default. What we're describing is narrower and, we think, more interesting: on the career pages of large multinational employers, for data science roles specifically, the disclosure rate isn't low. It's zero of 263, confirmed through both the structured salary fields these companies fill in and the free-text descriptions our new parser now reads.

How This Compares to Other Reports

A handful of "we analyzed N job postings" reports on data science salaries publish every year, and it's worth being upfront about how ours fits alongside them rather than pretending it exists in a vacuum. 365 Data Science's 2026 research, based on their own dataset, put overall disclosure at 37.8%. Ours comes in at 45.2% on a comparable structured-plus-extracted basis, most plausibly because our sample leans more heavily toward the US, where legal mandates push the number up. Neither figure is "more correct" than the other; they're measuring overlapping but not identical populations, pulled on different dates, with different source coverage. What neither of the other reports we could find does is break disclosure out by legal mandate status or run a same-country comparison against India, which is the part of this analysis we'd point to as new.

What This Actually Means if You're Job Hunting

  • If you're applying in California, New York, Washington, Massachusetts, Minnesota, or Illinois, you'll see a salary range on most listings you look at. Use that. Ranges are legally required to reflect a genuine good-faith estimate in these states, not a placeholder.
  • If you're applying in Texas, Georgia, or Pennsylvania, expect to ask, but don't assume silence means nothing to find. Read the actual job description text, not just whether the listing shows a salary badge, some job boards only check a structured field and miss ranges written directly into the description.
  • If you're applying in India, expect to negotiate blind, even at large, well-known multinational employers. Build your target range from outside data (Levels.fyi, Glassdoor, PayScale, peer conversations) before the first call, because the job posting itself almost certainly won't tell you.
  • Watch for globally-templated postings. A small number of listings we found carried a US salary range on a job based somewhere else entirely, a leftover from how some large employers template job descriptions across countries. If a number looks disconnected from the role's actual location, verify it before using it to set expectations.
  • We built browsing by disclosed salary range into finddatasciencejobs.com for exactly this reason: it's one filter that quietly tells you a lot about how a company treats candidates before you've had a single conversation with them.

Frequently Asked Questions

Which US states require salary ranges in job postings? As of 2026, 13 US jurisdictions require salary ranges to appear directly in job postings: California, Colorado, Washington, New York, Hawaii, Illinois, Minnesota, New Jersey, Vermont, Massachusetts, Maryland, Rhode Island, and Washington DC. Five of these took effect during 2025. Several other states require disclosure only upon request rather than in the posting itself, and more have pending legislation, so this list is worth rechecking closer to any hiring decision.

Is salary transparency required by law in India? No. India currently has no statutory requirement for employers to disclose salary ranges in job postings, which is consistent with what we found: zero of 263 India-based data science listings in our dataset disclosed a real salary, across 77 different employers, after correcting for a small templating bug that briefly made it look like three did.

Does the applicant tracking system a company uses affect whether I'll see a salary? Only if you're relying on a structured salary field. Workday, Workable, SmartRecruiters, and Greenhouse's public feed don't provide one, so we also check the job description text directly for a range, which is where a meaningful share of their disclosed salaries actually show up. Worth knowing on other job boards too: check the full listing, not just whether a salary badge appears.

Why would the same company disclose salary in one state but not another? Usually it doesn't, at least not as often as state-level averages suggest. Most of the gap in our data is a composition effect: companies that disclose everywhere as a default policy happen to be concentrated in certain states, and companies that rarely disclose are concentrated in others. Visa is the one clear exception in our data, disclosing meaningfully more often in law-mandated states than elsewhere, but it's the exception, not the rule.

How often will this data be updated? We're planning to refresh this analysis quarterly using live listings on finddatasciencejobs.com, checking both structured salary fields and job description text each time, so the gap between mandate and non-mandate states, and the US-India comparison, can be tracked on a consistent basis going forward.

about the author

Shaqeeq Khan is a growth marketer, designer, and startup builder based in Colombo, Sri Lanka. He is currently CMO and Co-Founder at Clarp.io and former Chief Growth Officer at Zyner, and has previously led marketing for Metana a web3 startup, scaling it from $0 to $1M ARR. His work sits at the intersection of growth, recruitment, and product — including experience with hiring systems and applicant-facing tools, alongside building products like AccentBridge and an early-stage voice AI MVP. He holds a BA (Hons) in Graphic Design from Falmouth University.