01
September: AI-skill demand is broader than AI/ML job titles
CompTIA reported 349,821 active US postings requiring AI-related skills in September 2026, up 32,327 from August. Dedicated AI/ML titles accounted for 13,636 postings. These measures describe different scopes: AI skills can be required in many jobs that do not carry an AI or machine-learning title.
CompTIA counted 272,040 new IT openings and 625,633 active IT postings. Software developers and engineers had 47,598 new openings, up 8.3 percent from August. These are postings, not hires.
In CompTIA’s employment analysis, tech-sector employment fell by about 10,350. Tech-worker unemployment was 3.3 percent, versus 4.2 percent nationally. Employment estimates and job-posting counts are different measures; monthly occupation data can be volatile and revised.
02
August: occupation growth and company cuts occurred together
CompTIA’s August 2026 analysis found tech-occupation employment increased by 86,000 while tech companies shed 14,700 jobs. AI-related postings surpassed 320,000, up 4.5 percent from July.
Tech occupations span employers throughout the economy. A staffing pullback at tech companies is not the same as a fall in demand for technical skills, and it does not establish that production AI engineers are readily available.
03
August AI/ML postings grew year over year, not quarter over quarter
Dice, using Lightcast data, reported AI and machine-learning tech postings were up 101 percent in August 2026 versus August 2025. This is an August year-over-year comparison, not a quarterly figure.
Do not combine this comparison with CompTIA’s broader AI-skill posting counts to calculate a Q3 total or growth rate. The sources measure different categories.
04
The broader September labor market remained weak
The BLS September 2026 Employment Situation, released 2 October 2026, reported payroll growth of 29,000 and unemployment of 4.2 percent. Information and professional and business services changed little. BLS did not publish an AI-occupation figure in that release.
Indeed Hiring Lab’s snapshot as of 18 September 2026 put the US Job Postings Index at 103.5, while tech postings remained depressed. This index is not an AI-jobs total, and this report does not state a September AI-mention share.
A weak overall payroll reading can coexist with competition for specific senior AI skills. Neither the BLS release nor the Indeed index resolves the availability of engineers for your particular production system.
05
Historical context, not Q3 2026 measurements
These earlier measures provide context only. They must not be relabeled as Q3 results or combined into a quarterly average.
- Indeed Hiring Lab, June 2026: AI-related postings were 5.9 percent of US postings. This is a June snapshot, not a Q3 or September AI share.
- Indeed Hiring Lab, through May 2026: 71 percent of the rise in software-development postings came from senior roles. This describes an earlier recovery period, not July through September.
- SHRM, July 2025 to June 2026: 28.5 percent of US IT and computer-science postings mentioned AI skills. This is a 12-month measure, not a Q3 figure.
- ZipRecruiter, June 2026 hiring-manager survey: 60 percent said AI had not changed headcount, 24 percent were hiring more and 16 percent fewer. These are survey responses from June, not Q3 payroll or posting counts.
06
What founders should do with these signals
Hire for production work rather than an AI label. Define whether the person must build agents, integrate models into a product, own evaluation or operate data pipelines. Make the reliability, security and delivery responsibilities explicit.
Expect competition for senior people who have owned working systems. The earlier Indeed evidence points to a senior-led software-posting recovery; it is not evidence that the Q3 market supplies every skill your team needs.
Do not read tech-company layoffs as a surplus of AI engineers. Scope the work, assess evidence of production ownership and distinguish employer-sector cuts from demand for specific occupations.
If you are considering a junior seat, be clear about the supervision and support it requires. Do not treat an urgent production role as an apprenticeship or assume AI removes the need for experienced oversight.