Software Engineering Job Market 2025–2030: The Forecast

The honest read on the software engineering job market between 2025 and 2030 is that two true statements look like a contradiction. The entry tier is contracting right now, and the profession is forecast to keep growing through the decade. Most takes pick one of those and call it the answer. This forecast holds both, traces each to a named source, and ends on a verdict rather than a hedge.

The reconciliation runs through a single distinction: headcount and job content move on different clocks. The number of developer positions is projected up. The work inside each position is being rewritten faster than the headcount changes. Read those two as one signal and the data turns to noise.

Why the Entry Tier Is Shrinking First

The sharpest signal in the recent data is age-specific, and it is negative. A November 2025 Stanford Digital Economy Lab study, "Canaries in the Coal Mine?" by Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen, used ADP payroll records covering millions of workers at tens of thousands of firms from 2021 through July 2025. It found employment for software developers aged 22 to 25 fell roughly 20% from its late-2022 peak, the moment generative AI tools went mainstream. Developers over 26 at the same firms stayed flat or grew.

The mechanism is not mysterious, and the authors are explicit about it. AI is strongest at exactly what early-career developers traditionally sell: textbook syntax, standard algorithms, the well-documented task with a known answer. Experience that AI cannot yet replicate, judgment about systems, tradeoffs, and failure modes, holds its value. The squeeze lands on the replaceable parts of the job, which happen to be concentrated at the bottom of the ladder.

Two design choices in the study make the finding harder to wave away. The decline did not appear in low-AI-exposure occupations such as health aides, which argues against a generic-recession explanation. And the pattern held across the most AI-exposed roles broadly, where early-career workers saw a 13% relative employment decline since late 2022. The youth effect is specific to AI exposure, not to the economy at large.

Verdict: the contraction is real, recent, and concentrated. It is an entry-level and code-typing phenomenon, not evidence that software engineering is in decline. Anyone citing the 20% number as proof the field is dying has read the headline and skipped the cohort.

The Label on the Job Decides the Trend

The Bureau of Labor Statistics splits a distinction the discourse usually flattens. Its 2023–2033 projection has software developer employment growing 17.9%, far above the 4% average for all occupations, adding roughly 327,900 jobs. In the same projection set, the narrower computer programmer category, the one closest to pure code-typing, declines 9.6%. Two roles a layperson would call "coding," moving in opposite directions, because they measure different work.

BLS is candid about why programmers fall. Its own outlook notes that companies are expected to use AI to automate repetitive programming tasks, with higher-skilled work shifting to software developers. The decline is not the disappearance of the activity; it is the activity migrating up the value chain into a role with a different title. The headline "AI is killing coding jobs" describes a reclassification more than an extinction.

The current market is louder than the projection and pointing the same way. CNN's April 2026 reporting ("The demise of software engineering jobs has been greatly exaggerated") documented developer openings growing as firms expect to ship more software now that AI lowers the floor on who can write code, raising demand for seasoned engineers to shape those products. IBM was tripling U.S. entry-level developer hiring on the logic that AI-augmented juniors can now take on work that once needed experienced staff. Listings were up about 30% over the year and more than 67,000 openings were tracked across major employers, even against 52,000 tech layoffs in the first quarter.

Verdict: "are developer jobs growing or declining" is the wrong question because it assumes one answer. Developers grow, programmers shrink, and the same automation drives both. The trend you find depends entirely on the occupation code you look up.

The 2030 Horizon Still Reads as Growth

The longest-range forecast in wide circulation does not treat software development as a sunset field. The World Economic Forum's Future of Jobs Report 2025 ranks software and application developers among the five fastest-growing roles to 2030 and fourth in percentage-growth terms, ahead of all but big data specialists, fintech engineers, and AI and machine learning specialists. That sits inside a projected net increase of 78 million jobs globally: 170 million created, 92 million displaced.

The same report attaches a number to how much the work changes underneath that growth. On average, employers expect 39% of workers' existing skill sets to be transformed or made outdated over the 2025 to 2030 window, with AI and big data the single fastest-rising skill demand. For a developer, the growth line and the obsolescence line describe one job, not two: more positions, different tasks inside them. A headcount forecast says nothing about whether today's day-to-day survives the decade.

The decade-out view from the buyer's seat agrees on direction and refuses the replacement framing. Gartner's survey of more than 700 CIOs, conducted in July 2025 and presented in November 2025, found that by 2030 CIOs expect 0% of IT work to be done by humans without AI, 75% by humans augmented with AI, and 25% by AI working alone. Augmentation is the modal expectation by a wide margin, with a real but minority autonomous slice. Nobody at the top of the budget is forecasting an empty engineering org.

A caution belongs next to the optimism. In a separate May 2025 Gartner survey of 506 technology leaders, 72% reported their organizations were breaking even or losing money on AI investments. Growth in demand for engineers is partly demand to make these investments finally pay, which is its own form of job security. The work exists because the automation is not yet reliable enough to leave alone.

Verdict: every multi-year forecast that names its method lands on growth for the role and transformation of the task. The disagreement in the data is about timing and tier, not about whether the profession has a 2030.

Headcount and Job Content Move at Different Speeds

The apparent contradiction dissolves once headcount and job content are read as separate variables. Headcount, the count of developer positions, is projected up by BLS and the WEF across the decade. Job content, the mix of tasks inside a position, is changing fast enough that 39% of skills are expected to turn over by 2030. A field can add jobs while gutting the daily work of each one, and software engineering is doing exactly that.

The near-term contraction and the long-term growth are the same transition viewed at different magnifications. Zoom in to 2023–2025 hiring and the entry tier is shrinking as employers stop paying juniors to do what AI now does. Zoom out to 2030 and total demand rises as AI-built software multiplies and someone has to architect, review, and run it. The juniors squeezed today are squeezed out of a task that is vanishing, not out of a profession that is shrinking.

This is why the most useful unit of analysis is the task, not the title. CNN's April 2026 reporting captured the shift in plain terms: developers are doing less routine coding and spending more time overseeing swarms of code-writing agents and designing software structure. The position survives because its center of gravity moves from producing code to directing and verifying it. The engineer who only typed is the one whose role contracts.

Verdict: there is no paradox, only two clocks. Headcount grows over the decade; job content is rewritten well inside it. Confusing the two produces every bad take in this category.

The Forecast Translated Into a Career Bet

For someone choosing whether to enter or stay in the field, the data points one direction with one condition attached. Software engineering remains a growth career to 2030 for anyone whose value sits above the line AI has already crossed. Below that line, the syntax recall and the standard-algorithm fluency that filled junior job descriptions, the floor is rising and the entry path is narrower than it was in 2021.

The practical move for new entrants is to skip past what AI already does well and arrive at the parts that hold their value: system design, code review at scale, debugging across services, and orchestrating agents rather than competing with them. The Stanford finding that experienced developers kept growing while juniors fell is a map, not just a warning. It shows which capabilities the market still pays a human premium for.

For employers, the contradiction in the data is a hiring trap. Cutting the entry tier because juniors are now slower than agents starves the pipeline that produces the senior engineers every forecast says will be in higher demand by 2030. IBM's decision to triple entry-level hiring while leaning on AI augmentation is the more defensible read of the same numbers: use AI to make juniors productive faster, not to stop hiring them.

Verdict: the field is a sound long-term bet with a narrowed front door. Enter above the automation line, build past it fast, and the 2030 demand forecast is on your side. Enter to do what an agent already does, and the near-term data is the warning.

The Single-Sentence Forecast

The defensible forecast for the software engineering job market from 2025 to 2030: the entry tier and the pure-coding roles contract now, the profession grows across the decade, and the reconciliation is that headcount rises while job content is rewritten around oversight and design. BLS projects 17.9% developer growth against a 9.6% programmer decline; Stanford measures a 20% entry-tier drop already; the WEF and Gartner both forecast growth and augmentation to 2030. Cite the cohort, name the occupation code, and the contradiction disappears.

For the full numerical picture behind this forecast, see the companion piece on AI software development statistics for 2026, and the section pillar on the future of software development for how the rest of the workflow is shifting.

Are software engineering jobs growing or declining?

Both, depending on which slice you measure. The U.S. Bureau of Labor Statistics projects software developer employment to grow 17.9% from 2023 to 2033, while the narrower computer programmer category declines 9.6% over the same window. The aggregate trend is growth; the entry tier and the pure code-typing roles are the parts under pressure. A Stanford payroll study found employment for developers aged 22 to 25 fell roughly 20% since late 2022, even as developers over 26 held steady or grew.

Will there be software engineering jobs in 2030?

Yes, and more of them than today. The World Economic Forum's Future of Jobs Report 2025 ranks software and application developers among the five fastest-growing roles to 2030, inside a net global increase of 78 million jobs. Gartner's November 2025 CIO survey expects 0% of IT work to be done by humans without AI by 2030, but 75% by humans augmented with AI and only 25% by AI alone. The job persists; its daily content is rewritten around oversight rather than typing.

How many software developers will there be in 2030?

BLS does not publish a single 2030 headcount, but its 2023–2033 projection adds roughly 327,900 software developer jobs to a base of about 1.8 million, implying well over 2 million by the early 2030s. The WEF places software developers among the largest absolute job gainers worldwide to 2030. The precise number depends on how the category is defined, since the line between developer, programmer, and AI-augmented technical worker is blurring.

Is software engineering a safe career?

Safe at the profession level, not at every entry point. The growth forecasts hold for experienced developers who own systems, tradeoffs, and judgment. The risk concentrates at the bottom of the ladder, where AI is strongest at exactly what juniors traditionally sold: textbook syntax and standard algorithms. The career remains durable for those who skip past what AI already does well and move toward architecture, review, and agent orchestration.

Which software engineering roles are growing?

Roles that involve designing systems, reviewing AI-generated output, and orchestrating coding agents are expanding, while routine line-by-line coding contracts. The WEF flags AI and machine learning specialists, big data specialists, and fintech engineers as the fastest-growing in percentage terms, with software and application developers fourth. CNN's April 2026 reporting noted demand rising for seasoned engineers to shape AI-built products, with IBM tripling U.S. entry-level developer hiring on the back of AI-augmented capability.