Top 5 Jobs in Government That Are Most at Risk from AI in San Diego - And How to Adapt

San Diego's public workforce is squarely in the path of an accelerating AI shift: governments “have access to tremendous amounts of data,” but adoption is uneven and often trails the private sector, so routine, form-driven roles like permit processing, records clerks, and benefits eligibility work become obvious beachheads for automation (SAS report on AI in government adoption and insights).

Studies show AI is already a reality in service delivery and internal operations, and practical pilots can cut the time spent on compliance and permit backlogs while leaving high-stakes decisions to humans; local examples include AI-assisted permit triage and case-management dashboards that free staff for in-person work (Neudesic analysis on AI speeding permits and casework).

For government workers seeking practical, on-the-job skills, the 15-week AI Essentials for Work 15-week bootcamp - prompt writing and AI tools for public sector teams teaches prompt writing and tool use to help public-sector teams redesign roles, not just replace them.

“This survey certainly indicates that AI isn't as far along in government as it is in virtually every other industry… there's a fairly high level of awareness of AI capabilities.” - Steve Bennett, director of the SAS global government practice

To pick the five San Diego government roles most exposed to automation, the methodology blended three practical lenses: (1) agency-level reporting and audits - including the California Department of Technology survey that found nearly 200 agencies claiming no high‑risk automated decisionmaking, a result some observers called “befuddling” (CalMatters analysis of California AI risk report); (2) the evolving legal and regulatory context that shapes what employers and municipalities can and cannot automate (new FEHA-aligned ADS rules and pending bills that would require notice and testing before AI-driven employment decisions) (Overview of California employment AI rules and regulations); and (3) occupation‑level risk and equity data - using automation risk frameworks and UCLA LPPI's breakdown of which California workers are concentrated in high‑automation jobs - to make sure selections reflect who will be affected most and why (UCLA LPPI analysis of automation risk for California Latinos).

That mixed approach prioritized routine, form‑driven tasks tied to large data sets or repeatable decisions (permit clerks, records staff, schedulers, eligibility processors) and weighted demographic exposure and legal vulnerability so the “so what?” is clear: job automation is not just technical risk, it's a policy and equity problem that demands proactive reskilling and safeguards.

“I only know what they report back up to us, because even if they have the contract… we don't know how or if they're using it, so we rely on those departments to accurately report that information up.” - Jonathan Porat, CTO, California Department of Technology

San Diego County administrative support specialists - those who juggle data entry, invoice tracking, document archiving, scheduling and routine correspondence - are squarely in AI's crosshairs because their day is packed with predictable, form‑driven work that automation handles well; Lapala's practical guide shows how tasks like archiving, invoice management and workflow digitization can be turned into low‑value automation wins that free staff for higher‑touch duties (Lapala guide to administrative task automations).

Modern toolsets automate email and calendar management, generate templates and update CRMs automatically, meaning many front‑desk cycles and repetitive approvals can be condensed or routed by software rather than by hand - see Lindy's roundup of time‑saving administrative automations for examples like inbox triage and database updates (Lindy AI administrative automations roundup).

For county managers worried about backlogs and morale, the calculus is simple: automate predictable bottlenecks (scheduling, reminders, FAQ responses) so human specialists can focus on exceptions, outreach and equity‑sensitive work - exactly the shift My AI Front Desk recommends for reclaiming hours lost to repetitive admin chores (My AI Front Desk automating administrative tasks guide).

San Diego Police Department records clerks handle a steady stream of predictable, sensitive tasks - selling crash and crime reports, conducting computer searches for criminal histories and impounds, registering sex, narcotics and arson offenders, fingerprinting the public, and entering stolen and recovered vehicle, property, pawn, gun, and boat records into state and national law‑enforcement systems - which makes them both indispensable and exposed to automation pressures; the official City class spec lays out a compact job footprint and pay band for this work (City of San Diego Police Records Clerk class specification and pay band).

Because much of the role is standardized data entry, redaction and record‑retrieval logic, it's a prime candidate for LLM‑augmented search, automated redaction and workflow routing - changes that can speed service but demand careful privacy and release‑of‑information safeguards reflected in local policy.

For managers and affected staff, the practical takeaway is clear: the clerical core is routine enough to automate, yet sensitive enough that role redesign must pair efficiency with stronger controls (comprehensive police records clerk job description and duties guide).

City of San Diego permit technicians - whether titled Development Services Permit Technician or a county Land Use Aide - sit at the junction of public-facing intake, technical plan checks, and repetitive permit‑tracking work, which makes them both essential and exposed to automation: the class spec shows technicians are tasked with assisting customers with plan check filing, processing fees, routing plans, and

reviewing and evaluating plans for completeness

while using permit tracking systems (Development Services Permit Technician class specification and duties); county job descriptions underscore similar intake, zoning and building review duties for entry‑level Land Use Aides (San Diego County Land Use Aide job description and responsibilities).

Because many steps - no‑plan permit issuance, fee calculation, basic completeness checks and digital plan routing - are rule‑driven, modest automation can speed throughput and free staff for complex consultations, but the

so what?

is clear: when permit queues and developer deadlines collide, preserving human oversight at technical touchpoints is crucial to protect safety, equity, and code compliance as the city moves more services online (see the San Diego Development Services permits and approvals portal for scope and timelines) (San Diego Development Services – Permits & Approvals portal and timelines).

San Diego County Eligibility Specialists who process CalFresh and Medi‑Cal applications spend their days scheduling interviews, guiding applicants through sensitive verification, documenting income and residency, and arranging home visits when paperwork is inconclusive - tasks tightly governed by rules that both open the door to automation and demand careful human oversight.

California requires counties to schedule initial interviews promptly so eligible households get a chance to participate within 30 days, and many counties now use on‑demand phone interviews or telephonic scheduling to speed intake (CalFresh interview process and interview scheduling guidance); BenefitsCal's interview tips and multilingual resources underscore that interviews are often telephonic, that applicants have enforceable rights (interpreters, privacy), and that workers must help complete missing application information (BenefitsCal interview tips and applicant rights).

Routine, rule‑driven pieces - appointment booking, preliminary document checks, automated reminders and electronic verification - are ripe for workflow automation, but anything involving interpreter services, privacy‑sensitive disclosures, helping a household complete an application, or scheduling and conducting a home visit resists full automation; the policy tradeoffs matter when a missed interview can be rescheduled within the 30‑day window but a second missed interview may lead to denial unless the household requests otherwise (Preparing for your CalFresh interview and applicant obligations).

So what?

The answer is plain: automate the predictable, protect the procedural safeguards, and preserve specialist time for high‑touch fairness work.

San Diego's Metropolitan Transit System schedulers - whose remit centers on routing, timetable decisions and keeping vehicles moving on time - are squarely in AI's sights because the same tools that optimize commercial supply chains can trim transit costs and emissions by finding more efficient routes and delivery schedules (AI-driven logistics and route optimization - University of San Diego PCE).

That means routine schedule adjustments, headway planning, and reactive rerouting during traffic or incidents are increasingly automatable, and agencies that buy these systems need tested procurement playbooks and transparency clauses so efficiency gains don't come at the expense of equity or rider privacy (Procurement tips for city buyers to govern AI in transit).

At the same time, LLM‑augmented tools can strengthen resilience - accelerating incident triage or generating operations playbooks for IT and dispatch teams - if paired with bias and privacy safeguards from the outset (Privacy and bias safeguards for AI in public transit operations).

The bottom line: smart schedulers will be those who can work alongside algorithms that shave minutes from trip times and, in doing so, help reduce vehicle miles and local emissions while protecting riders.

California's practical path forward pairs paid training, registered apprenticeships, and easily accessible state resources so public workers aren't left behind: tap into the Employment Training Panel's employer-driven grants to underwrite incumbent‑worker upskilling (Employment Training Panel (ETP) program guidance and funding), pilot an “earn‑while‑you‑learn” Registered Apprenticeship Program so entry‑level hires and mid‑career clerks gain credentials while on the job (ILG Bridge Registered Apprenticeship Program (RAP) initiative), and connect affected staff to no‑cost retraining and placement services through the EDD and local America's Job Centers (EDD Resources for Job Seekers and America's Job Centers).

Pairing these investments with clear procurement rules, bias and privacy safeguards, and role‑redesign that reserves human judgment for exceptions turns automation from a job threat into a workforce upgrade - imagine a records clerk who trades repetitive data entry for supervising LLM‑assisted redaction workflows and handling the toughest privacy calls, all while retaining pay and benefits during retraining.

The bottom line for California: AI is reshaping routine work faster than many expect, so preparing San Diego's public workforce means pairing sharper procurement rules and privacy safeguards with real, paid upskilling that changes job content - not just headcount.

National-level research warns that roughly 30% of U.S. jobs could be automated by 2030 and that a majority of roles will see significant task change, so local leaders should treat automation as a workforce transition (see the National University roundup on AI job trends).

Local evidence shows AI demand climbing in San Diego (1.96% of job postings referenced AI in 2022), which makes region-specific training and apprenticeships essential (UC San Diego Extended Studies analysis).

Practical steps include transparent procurement, funded incumbent training, and short, work‑focused courses that teach prompt writing and tool use so staff can supervise AI instead of being replaced - one accessible option is the 15‑week AI Essentials for Work bootcamp: practical AI skills for non-technical public employees, which teaches AI tools, prompt design, and job‑based AI skills for non‑technical public employees.