Stack Overflow Developer Survey 2026
Organization's measures on AI cost and usage by age
In the Under 18 years old group, 46% selected “None of the above”; in the 35-44 years old group, 14% did. Of the 10 answers tested, “None of the above” differs most by age, a small association: χ²(7, N = 11468) = 221.38, Holm-adjusted p < .001, V = .14. Respondents could select more than one answer, so the percentages add to more than 100%. These are unweighted respondents, not population estimates.
How results are checked
Named statistical review pending. Selected automated reference cases are published on the validation page.The table
The table
Question asked: Which, if any, of the following has your organization done related to AI tool or model usage? Select all that apply.
Respondents could select more than one answer. Shown by age, as the percentage of each group's respondents who selected each answer, with the number of respondents beneath. A respondent is counted in every row they selected, so a column adds to more than 100%.
| Organization's measures on AI cost and usage | Under 18 years old (A) | 18-24 years old (B) | 25-34 years old (C) | 35-44 years old (D) | 45-54 years old (E) | 55-64 years old (F) | 65 years or older (G) | Prefer not to say (H) | All |
|---|---|---|---|---|---|---|---|---|---|
| Set limits on AI tool or model usage or spending | 17.1%7 | 25.9%209 | 38.6%BG1,332 | 41.8%ABFG1,593 | 39.4%BG815 | 34.5%BG335 | 22.2%58 | 33.3%18 | 38.1%4,367 |
| Reviewed internal usage or spending data for AI tools or models | 9.8%4 | 21.0%170 | 27.5%BFG951 | 32.3%BCFGa1,229 | 29.5%BFG611 | 22.3%216 | 17.6%46 | 24.1%13 | 28.3%3,240 |
| Approved different AI tools or models for different types of work | 12.2%5 | 20.9%169 | 25.9%Gb893 | 30.5%BCGf1,161 | 28.7%BG594 | 25.5%g247 | 16.5%43 | 18.5%10 | 27.2%3,122 |
| Set limits on who can use certain AI tools or models | 12.2%5 | 19.1%154 | 24.8%BG856 | 27.5%BG1,046 | 27.5%BG570 | 23.5%G228 | 13.0%34 | 14.8%8 | 25.3%2,901 |
| Increased spending on AI tools or models because results justified it | 7.3%3 | 16.2%131 | 18.4%635 | 23.0%BC875 | 22.4%BC463 | 20.7%201 | 18.8%49 | 16.7%9 | 20.6%2,366 |
| Not sure | 24.4%10 | 18.8%152 | 17.5%606 | 17.7%676 | 18.1%374 | 17.6%171 | 16.5%43 | 20.4%11 | 17.8%2,043 |
| None of the above | 46.3%CDEFb19 | 25.1%CDE203 | 16.8%D581 | 13.8%527 | 16.8%d347 | 23.6%CDE229 | 40.2%BCDEF105 | 35.2%CDE19 | 17.7%2,030 |
| Compared buying an AI tool with building something internally | 4.9%2 | 11.1%Gf90 | 12.9%FGe447 | 11.9%FG453 | 10.3%G213 | 7.0%68 | 4.2%11 | 11.1%6 | 11.2%1,290 |
| Stopped using or reduced spending on other software tools while increasing investment in AI tools or models | 4.9%2 | 7.9%G64 | 9.0%Gf311 | 10.2%EFG390 | 7.5%G156 | 6.0%58 | 2.3%6 | 9.3%5 | 8.7%992 |
| Negotiated shorter commitments or the ability to switch AI tools or models over time | 4.9%2 | 3.6%29 | 3.0%104 | 4.0%Fe154 | 2.6%53 | 1.5%15 | 1.5%4 | 5.6%3 | 3.2%364 |
| Respondents | 41 | 808 | 3,453 | 3,810 | 2,071 | 970 | 261 | 54 | 11,468 |
A letter beside a percentage names the columns it is significantly higher than (uppercase 95%, lowercase 90%; pairwise column tests, Bonferroni-corrected). 19,435 of 30,903 respondents have no answer to one of the two questions (not asked or not answered) and are left out.
Analysis
Results
Chi-square test of “None of the above”
χ² = 221.38
df = 7, Holm-adjusted p < .001
Effect size
Cramér's V = 0.139
a small association
The strongest answer, APA-style: χ²(7, N = 11468) = 221.38, Holm-adjusted p < .001, V = .14. Each answer is tested on its own, as the respondents who selected it against those who did not, across the 8 groups (N = 11,468). The p-values are Holm-adjusted for the 10 answers tested; 9 of them differ by age at the 0.05 level.
| Answer | Selected | χ² (df = 7) | Adjusted p | Cramér's V |
|---|---|---|---|---|
| Set limits on AI tool or model usage or spending | 38.1% | 116.59 | < .001 | 0.101 |
| Reviewed internal usage or spending data for AI tools or models | 28.3% | 92.41 | < .001 | 0.090 |
| Approved different AI tools or models for different types of work | 27.2% | 65.45 | < .001 | 0.076 |
| Set limits on who can use certain AI tools or models | 25.3% | 61.22 | < .001 | 0.073 |
| Increased spending on AI tools or models because results justified it | 20.6% | 42.18 | < .001 | 0.061 |
| Not sure | 17.8% | 2.61 | .919 | 0.015 |
| None of the above | 17.7% | 221.38 | < .001 | 0.139 |
| Compared buying an AI tool with building something internally | 11.2% | 45.52 | < .001 | 0.063 |
| Stopped using or reduced spending on other software tools while increasing investment in AI tools or models | 8.7% | 39.35 | < .001 | 0.059 |
| Negotiated shorter commitments or the ability to switch AI tools or models over time | 3.2% | 24.67 | .002 | 0.046 |
How to read it
No single chi-square test covers this table. A respondent appears in every row they selected, so the rows are not separate groups of people, and a test of the whole table would count the same person several times. Each answer is tested on its own instead: did the share who selected it differ between the groups? Testing 10 answers gives 10 chances of a false alarm, so the p-values are Holm-adjusted to keep the chance of any false alarm in this table at 0.05.
For “None of the above” the test rejects independence at the 0.05 level (Holm-adjusted p < .001): differences this large between the groups are unlikely if selecting it did not depend on age. Cramér's V of 0.139 is in the small range, which is real but modest: knowing someone's group tells you only a little about their answer.
With 11,468 respondents, even small differences reach statistical significance. Read the percentages and the effect size before the p-value.
Limits
- The survey has no weights. These tables are unweighted counts of the people who answered, so they describe the respondents, not developers in general.
- The test treats respondents as a simple random sample. They are people who chose to answer a survey Stack Overflow promoted, not a random sample of developers, so a result holds for these respondents and may not hold beyond them.
- The adjustment covers the answers of this question by this group only, not the other tables in the library. The headline answer is the strongest of those tested, so its effect size is larger than a typical answer's.
- A difference between groups is not evidence that group membership causes the answer.
Next steps
Work with this table
Open the test of “None of the above” in the calculator
The link fills in the respondents who selected this answer and those who did not, by group. The whole table does not open there: its rows overlap, and the calculator's test needs rows that are separate people.
Open this question in the full file
The workspace opens all 30,903 respondents in your browser with this question selected. The survey has no weights, so the table there is unweighted too. Change the group, add a filter or export from there.
Open in the workspace →Run this on your own data — free, no signup
Upload a CSV, Excel, SPSS, Stata, SAS, R, Parquet or JSON file. Everything runs in your browser; your file never leaves your device.
Open the workspace →The same question by other groups
Other questions that differ by age
- Other work, school or life responsibilities by age
- How they learned to code in the past year by age
- Activities in a typical work week by age
- How they learned to code for AI in the past year by age
- Programming languages: Worked with in past year by age
- What prevents taking part in technical communities by age
Source
Stack Overflow. 2026 Developer Survey, https://survey.stackoverflow.co/2026, licensed under the ODbL 1.0. Tabulated by crosstabs.com. Variables AIToken (Organization's measures on AI cost and usage) and Age. Stack Overflow licenses the survey data under the Open Database License (ODbL) 1.0 and its contents under the Database Contents License (DbCL) 1.0: credit Stack Overflow, keep the licence notice, and share any database made from the data under the ODbL. These tables and their downloads are offered on those terms.