IT Asset Management Group (IT-AMG) conducted a study that utilized Google's Gemini 3.5 Flash model to identify occupations most likely to be automated by artificial intelligence. The model identified data entry clerks, telemarketers, Tier 1 customer service representatives, bookkeepers, proofreaders, paralegals focusing on document review, commodity content writers, market research analysts, retail cashiers, and commercial translators as jobs most vulnerable to automation.

The model stated that data entry jobs involve repetitive tasks, such as transferring text and numbers between documents and databases, which makes them relatively easy to automate. For telemarketing, AI-generated voices and automated calling systems can follow scripts at lower costs than human teams. Gemini also noted that large language models can answer common customer questions and sort support requests using company knowledge bases in Tier 1 customer service roles.

Regarding bookkeepers, Gemini stated that automation software increasingly handles routine tasks like reconciling ledgers and organizing expenses. For proofreaders, grammar and editing tools can quickly identify spelling, punctuation, and style issues, reducing the need for manual review. AI can rapidly scan contracts, legal filings, and case documents, assisting paralegals by reducing the amount of manual review.

For commodity content writers, basic blog posts, product descriptions, and advertising copy can often be generated quickly with AI tools. Gemini indicated that AI can collect data, analyze customer sentiment, and summarize findings for market research analysts, with humans continuing to make strategic decisions. Self-checkout systems and mobile payment technology reduce demand for traditional retail cashier roles. For commercial translators, AI language models are becoming more accurate for business and technical documents, though human editors are often still needed for quality control.

Richy George, chief revenue officer at IT-AMG, said, "What's interesting is how fast the middle layer of corporate data management is shrinking." George added, "Data entry ranking top shows that pure information transfer may no longer be a viable human career path."