In 2026, the ability to google was replaced by the ability to prompt.
While many industries are still just starting to embrace artificial intelligence, the tech industry has moved to the next stage: companies are monitoring how actively employees are using AI tools and are ready to coerce those who resist.
From small startups to giants like Amazon, Google, and Meta, employers are measuring AI use, hoping for productivity gains, and in some cases, even factoring results into performance assessments. For some positions, candidates without a solid grasp of AI tools are not even considered.
"We use both carrots and sticks," says Seth Besmertnik, CEO of Conductor, a 300-person marketing company. "The only way to build a successful company is to achieve a high level of competence across the entire workforce." Since the beginning of the year, Besmertnik has included AI proficiency in the company-wide performance appraisal system. Each employee receives a rating from one to five points, with fives awarded to those who create systems that improve their colleagues' work processes. Additionally, the company has instituted an award: the creator of the most effective AI process receives a sabbatical stipend of several thousand dollars.
Besmertnik also refuses to hire people without AI skills. During interviews, candidates are asked to solve test problems using AI, explain their choice of tool, demonstrate prompts, and explain how their approach would have changed six months ago. Two dedicated interviewers evaluate AI skills.
The scale of the phenomenon is confirmed by the numbers. According to consulting firm Section, by October of last year, 42% of tech workers said management expected AI to be used in their daily work—up from 32% just eight months earlier. Nearly half of telecom and IT companies are already seeing positive returns on their investments in generative AI, compared to an average of 35% across all industries, according to a joint study by the Wharton School and GBK Collective.
However, even in an industry that sets trends in the labor market, AI adoption is not going smoothly. Tech professionals share the same feelings as everyone else: skepticism about real time savings and anxiety over their own CEOs' announcements of workforce reductions thanks to automation. "Anxiety is growing in the tech sector," notes Jeremy Korst, co-author of the Wharton report. "Do we really think employees will embrace a tool they believe will cost them their jobs?"
Meanwhile, major IT companies are actively monitoring the use of AI within their organizations. At Amazon Web Services, developer managers see a dashboard with data on engineers' performance with AI tools. While these metrics don't formally influence performance reviews, managers do consider how actively an individual is involved in AI when deciding on promotions . Google, for the first time this year, began factoring AI use into performance reviews for some engineers. Teams and managers have the right to evaluate AI skills based on the employee's role, although such assessments are not yet mandatory.
At Meta*, a new performance appraisal system takes into account work with AI and can count how many lines of code an engineer has written using AI assistants . At Microsoft, managers include questions about AI in performance reviews—employees are required to specifically describe how they use AI tools in their work processes.
Salesforce went even further: at the end of last year, an AI proficiency tracker was added to its internal dashboard. While there are no formal metrics for assessment, tool usage and performance are closely linked. "If you're not using AI, you're probably underperforming," a company representative summarizes. "We measure adoption accurately," confirms Joe Inzerillo, Salesforce's president of enterprise and AI technologies. "We're obsessed with numbers." Salesforce's philosophy over the past year has become one of experimentation and iteration: the best tools naturally rise to the top, and when a critical mass of employees begins using a particular solution, it's time for mandatory implementation. Already, applying for paid leave in Salesforce is only possible through an AI agent, and most self-assessments and performance reviews are conducted with the AI assistant.
The pressure is particularly intense within large tech companies, which have invested billions in developing AI products. "Microsoft, Amazon, and Google are spending huge amounts of money on developing AI tools," says future of work consultant Brian Elliott. "If they can't get their own employees to use the products effectively, selling the solutions to customers will be much more difficult."
Andrew Anagnost, CEO of software developer Autodesk, admits that access was the main barrier to adoption: some coding tools, including Cursor, were initially blocked by corporate policy, and employees used them secretly. Instead of overwhelming people with tools, Autodesk focused on specific workflows that AI could improve. However, a small number of stubborn AI resisters will always remain. "These people are unlikely to last long," Anagnost warns.
While many industries are still just starting to embrace artificial intelligence, the tech industry has moved to the next stage: companies are monitoring how actively employees are using AI tools and are ready to coerce those who resist.
From small startups to giants like Amazon, Google, and Meta, employers are measuring AI use, hoping for productivity gains, and in some cases, even factoring results into performance assessments. For some positions, candidates without a solid grasp of AI tools are not even considered.
"We use both carrots and sticks," says Seth Besmertnik, CEO of Conductor, a 300-person marketing company. "The only way to build a successful company is to achieve a high level of competence across the entire workforce." Since the beginning of the year, Besmertnik has included AI proficiency in the company-wide performance appraisal system. Each employee receives a rating from one to five points, with fives awarded to those who create systems that improve their colleagues' work processes. Additionally, the company has instituted an award: the creator of the most effective AI process receives a sabbatical stipend of several thousand dollars.
Besmertnik also refuses to hire people without AI skills. During interviews, candidates are asked to solve test problems using AI, explain their choice of tool, demonstrate prompts, and explain how their approach would have changed six months ago. Two dedicated interviewers evaluate AI skills.
The scale of the phenomenon is confirmed by the numbers. According to consulting firm Section, by October of last year, 42% of tech workers said management expected AI to be used in their daily work—up from 32% just eight months earlier. Nearly half of telecom and IT companies are already seeing positive returns on their investments in generative AI, compared to an average of 35% across all industries, according to a joint study by the Wharton School and GBK Collective.
However, even in an industry that sets trends in the labor market, AI adoption is not going smoothly. Tech professionals share the same feelings as everyone else: skepticism about real time savings and anxiety over their own CEOs' announcements of workforce reductions thanks to automation. "Anxiety is growing in the tech sector," notes Jeremy Korst, co-author of the Wharton report. "Do we really think employees will embrace a tool they believe will cost them their jobs?"
Meanwhile, major IT companies are actively monitoring the use of AI within their organizations. At Amazon Web Services, developer managers see a dashboard with data on engineers' performance with AI tools. While these metrics don't formally influence performance reviews, managers do consider how actively an individual is involved in AI when deciding on promotions . Google, for the first time this year, began factoring AI use into performance reviews for some engineers. Teams and managers have the right to evaluate AI skills based on the employee's role, although such assessments are not yet mandatory.
At Meta*, a new performance appraisal system takes into account work with AI and can count how many lines of code an engineer has written using AI assistants . At Microsoft, managers include questions about AI in performance reviews—employees are required to specifically describe how they use AI tools in their work processes.
Salesforce went even further: at the end of last year, an AI proficiency tracker was added to its internal dashboard. While there are no formal metrics for assessment, tool usage and performance are closely linked. "If you're not using AI, you're probably underperforming," a company representative summarizes. "We measure adoption accurately," confirms Joe Inzerillo, Salesforce's president of enterprise and AI technologies. "We're obsessed with numbers." Salesforce's philosophy over the past year has become one of experimentation and iteration: the best tools naturally rise to the top, and when a critical mass of employees begins using a particular solution, it's time for mandatory implementation. Already, applying for paid leave in Salesforce is only possible through an AI agent, and most self-assessments and performance reviews are conducted with the AI assistant.
The pressure is particularly intense within large tech companies, which have invested billions in developing AI products. "Microsoft, Amazon, and Google are spending huge amounts of money on developing AI tools," says future of work consultant Brian Elliott. "If they can't get their own employees to use the products effectively, selling the solutions to customers will be much more difficult."
Andrew Anagnost, CEO of software developer Autodesk, admits that access was the main barrier to adoption: some coding tools, including Cursor, were initially blocked by corporate policy, and employees used them secretly. Instead of overwhelming people with tools, Autodesk focused on specific workflows that AI could improve. However, a small number of stubborn AI resisters will always remain. "These people are unlikely to last long," Anagnost warns.
