Key Takeaway
The new wave of automation is not eliminating occupations all at once. It is first changing tasks that are rule-based, digitally transferable and easy to split into small assignments. That is why the impact of generative artificial intelligence is appearing early in data entry, basic design, routine copywriting, parts of translation, bookkeeping support, first-line customer service and some entry-level coding tasks. A Financial Times analysis published on 14 September 2026 uses the gig economy to illustrate the shift: the easier a task is to describe and hand to software, the faster the market price of human-only execution can come under pressure.
For Georgia, the immediate risk is less a sudden wave of mass unemployment than a change in the structure of work. Routine assignments may become scarcer, entry-level competition may intensify and prices may fall for services whose output can be standardised. At the same time, demand should rise for workers who use technology while adding what automation does not easily replace: accountability, client context, local knowledge, human interaction, judgment and quality control.
Why gig and freelance work changes first
The Financial Times describes pressure on both sides of platform work. In digital services, simple graphic design, copywriting, data processing and other remote tasks increasingly compete with generative AI. Platforms that built their businesses by matching small clients with freelancers are trying to move toward more complex, higher-value services. In physical gig work, the shift is slower but visible: robotaxis and automated delivery could eventually affect ride-hailing and courier work as regulation and infrastructure permit.
The economics are straightforward. Gig work is often already broken into discrete tasks. A client does not buy a permanent job; it buys an output – a short text, a logo, a cleaned spreadsheet, a video description or a small software component. If software can produce a comparable first version faster and at lower cost, the barrier to substitution is low.
This is why automation usually reaches a profession through its most standardised tasks before it reaches the profession itself. A lawyer may remain indispensable while first-pass contract summaries are automated. An accountant may stay essential while manual data transfer declines. A marketing specialist may become more valuable by using AI for drafts and spending more time on strategy and client decisions.
What the international evidence shows
The International Labour Organization’s 2025 update estimates that roughly one in four workers worldwide is in an occupation with some degree of exposure to generative AI. Its conclusion is more cautious than a simple job-loss narrative: transformation is more likely than full replacement. Clerical occupations remain the most exposed, including data-entry clerks, typists, accounting and bookkeeping clerks and administrative secretaries.
In June 2026, the ILO reviewed a growing body of empirical evidence. It found that productivity gains and time savings from AI are real but uneven. Large-scale displacement remains limited so far. More immediate concerns include weaker entry-level opportunities, especially for younger workers, rising inequality and changes in work organisation that can affect autonomy and job quality.
This distinction matters. Inside a firm, automation may function as a productivity tool for an employee. In an open freelance market, the same technology may reduce the number of paid tasks available to a worker who sells only one standardised service.
Why Georgia is exposed in a different way
Georgia’s labour market is both protected from and exposed to automation. A large share of economic activity still depends on physical presence, personal interaction, local language and on-site work. Retail, construction, tourism, transport, agriculture and many small businesses cannot be replaced by software alone.
At the same time, a significant share of employment is outside the classic salaried model. According to Georgia’s National Statistics Office, Geostat, 1.365 million people were employed in the second quarter of 2026 and employees accounted for 69% of total employment. Self-employment therefore represented roughly 31%. According to calculations by BTU researchers, that is approximately 423 thousand people. This broad category is not the same as platform or freelance work, and it should not be treated as a proxy for the gig economy. It does, however, show that a large group of workers in Georgia earns income outside a standard fixed-wage relationship.
The unemployment rate was 13.8% in the same quarter, while the employment rate stood at 46.0%. In such a labour market, the significance of automation is not only how many existing employees lose jobs. It is also whether younger and less experienced workers can still find the first paid tasks that allow them to build a career.
Which jobs are closest to change
In Georgia, early exposure is likely to appear where the output is fully digital and basic quality control is relatively easy. This includes data entry and processing, routine office administration, standard copywriting, first-pass translation, repetitive social-media content, simple visual design, first-line customer support, document summarisation and some small coding tasks.
Geostat’s 2024 occupational distribution shows about 64,000 clerical support workers in Georgia – 38.4 thousand women and 25.6 thousand men. According to calculations by BTU researchers, the group represented about 4.6% of total employment in 2024, and women accounted for 60% of the group. This does not mean that a specific number of those jobs will disappear. The ILO framework also measures task exposure, not the probability of dismissal. But clerical support is the Georgian occupational group most directly aligned with the tasks that international evidence identifies as highly exposed.
The next area is digital creative and professional work. Basic text, first-stage visual design, routine video support and parts of software development are increasingly easy to automate. The important distinction is task complexity. The more generic the assignment, the more it can face price pressure. The more it requires industry knowledge, client-specific understanding, legal or commercial responsibility and an original decision, the more difficult full substitution becomes.
Georgia’s digital readiness can accelerate adoption
Geostat’s 2025 enterprise ICT survey found that 94.9% of enterprises had internet access. It also found that 35.5% of employees used portable computers or smartphones provided by their enterprise for business internet access. The basic connectivity required for rapid adoption is therefore widely present.
Digital maturity is much less uniform. Only 15.3% of enterprises reported using a webpage or website. A cautious interpretation is that many Georgian firms are connected, but the depth of digital business processes varies significantly. Automation is therefore likely to spread at different speeds – faster in banking, consulting, technology, marketing and other digitally intensive services, and more slowly where value is created mainly through physical operations.
The entry-level problem
One of the less visible labour-market risks is that many professionals learn through simple tasks. Junior designers make basic visuals. Legal assistants summarise documents. New programmers fix small pieces of code. Junior marketers draft standard copy. If those tasks are automated, the question becomes: where does a beginner acquire the experience needed to perform more complex work later?
The social risk of automation may therefore be not only fewer jobs, but a narrower first rung on the career ladder. For Georgia, where youth unemployment has historically been higher than the overall rate, this deserves particular attention. For companies, the same problem can return later as a shortage of experienced staff if junior development pipelines disappear.
How competitive advantage changes
Under automation, value shifts from task execution toward judgment. A freelancer who only produces generic text can face strong price pressure. A professional who researches the client’s market, asks better questions, builds strategy, verifies facts, understands Georgian context and takes responsibility for the final result provides a service that is far less standardised.
The same is true inside companies. Employee value will increasingly depend on the ability to use technology, verify output, detect errors and frame the right business problem. AI literacy will not remain a stand-alone specialist skill. In many occupations it will become a standard work tool, much as spreadsheets, search engines and online communication already have.
BTU Researchers’ Assessment
According to an assessment by BTU researchers, Georgia’s near-term automation effect is more likely to be a redistribution of tasks inside jobs and a repricing of freelance assignments than a simultaneous wave of large-scale layoffs. The strongest pressure will fall on low-value, repetitive and easily measurable digital work. Work that requires physical presence, trust, human interaction, deep knowledge of Georgian language and context, regulatory responsibility or complex judgment should be more resilient.
For Georgia, the right response is therefore not to resist technology. The priority is to move workers from simple execution toward higher-value use: people need not only to operate AI tools but also to evaluate outputs, protect data, define business problems correctly and take responsibility for final quality.
What business and policy should focus on
For firms, the first step should be mapping tasks rather than simply identifying jobs to automate. Companies should distinguish repetitive work that can be accelerated from decisions where human judgment is critical. This can raise productivity without discarding important organisational knowledge.
The second issue is junior development. If basic work is automated, firms need deliberate alternatives for learning: supervised assignments, quality control, client interaction and verification of AI-generated output. Otherwise, short-term savings can create a long-term talent shortage.
For government and education, the priority is faster curriculum renewal. The answer should not be coding alone. Workers need a combination of digital literacy, analytical reasoning, communication, domain expertise and responsible AI use. Accessible retraining is especially important for self-employed and income-insecure workers because they are less protected by employer-funded learning systems.
Conclusion
Automation is changing work, but its first effect is not the disappearance of whole professions. It first removes or cheapens tasks that are easy to specify, transfer and verify. That is why freelancers, clerical support workers, providers of simple digital services and people at the beginning of their careers can feel the change first.
Georgia’s challenge is to ensure that technological progress does not narrow labour-market opportunity. Companies need to use automation to extend worker capability rather than only to cut cost. Education needs to move people from routine execution toward judgment, accountability and complex problem-solving.
The likely winners will not be workers who somehow avoid AI entirely. There will be fewer such roles. The advantage will belong to people who combine technology with expertise and move toward the parts of work where human context and responsibility remain central.
Data and Main Sources
Financial Times, “Automation is coming for the gig economy”, 14 September 2026.
https://www.ft.com/content/a2c33eed-7aef-4e68-9252-3a1dddf61b5a
National Statistics Office of Georgia (Geostat), Indicators of the Labour Force – II Quarter 2026.
https://www.geostat.ge/media/82084/Indicators-of-the-Labour-Force-%28Employment-and-Unemployment%29—II-Quarter-2026.pdf
Geostat, Employment and Unemployment Data Module.
https://www.geostat.ge/en/modules/categories/683/Employment-Unemployment
Geostat, Women and Men in Georgia 2025 – occupational employment data for 2024.
https://www.geostat.ge/media/74964/WOMEN-AND-MEN—2025.pdf
Geostat, Use of Information-Communication Technologies in Enterprises – 2025.
https://www.geostat.ge/media/79412/Use-of-Information-Communication-Technologies-in-Enterprises—2025.pdf
International Labour Organization, Generative AI and Jobs: A 2025 Update.
https://www.ilo.org/publications/generative-ai-and-jobs-2025-update
International Labour Organization, The Impact of GenAI on Jobs, Productivity and Work Organization: A Review of the Empirical Evidence, 2026.
https://www.ilo.org/publications/impact-genai-jobs-productivity-and-work-organization-review-empirical
Prepared by the academic team of Business and Technology University and the BTUAI Research Team, Tbilisi, Georgia.