Meta Revises Employee Performance Review Criteria While Testing Autonomous Hatch AI Tool Meta has updated its performance review guidelines to eliminate strict AI usage metrics while encouraging staff to test its new autonomous Hatch AI agent. Technology giant Meta has substantially restructured its internal workforce appraisal system by decoupling employee performance reviews from mandatory artificial intelligence adoption metrics. Under revised guidance distributed internally this week, the company eliminated review criteria that tied job evaluation ratings to heavy utilization of AI applications or specific designations such as 'AI Native'. Instead, internal evaluators will measure staff members strictly on their tangible business contributions and overall impact, regardless of whether those results were generated using artificial intelligence tools or traditional operational methods. Simultaneously, Meta has begun encouraging employees to test its newest experimental autonomous AI system, designated as Hatch, prior to an anticipated public release. Revised Performance Review Guidelines and Ending Usage Tracking The updated appraisal instructions replace former performance criteria focused on 'usage of AI' and 'AI Native' classifications with broad language clarifying that business outcomes can be supported by AI or other operational means. This policy adjustment restores Meta's foundational focus on evaluating employees by the substance of their work rather than the specific software mechanisms used to achieve it. Staff members noted that the modification subtly yet meaningfully relieves pressure to deploy AI tools in workplace situations where their application does not make practical sense. Addressing the policy adjustments, Meta spokesperson Tracy Clayton emphasized that the newly issued instructions reflect the company's long-standing philosophy of rating employees based on their actual contributions. He clarified that labels including 'AI Native' were never incorporated as formal metrics during performance appraisals. Additionally, engineering teams across the organization were informed this week that management will not utilize AI adoption dashboards or token consumption metrics to assess employee performance or business impact. Approximately one year ago, Meta announced that worker evaluations would incorporate 'AI-driven impact', assessing the degree to which employees integrated chatbots and automated software agents into their daily tasks. Workforce members who logged high usage metrics were systematically tagged with classifications such as 'AI Native', 'AI First', or 'AI Enabled'. That initial policy framework spawned significant legal friction. A lawsuit filed in July by approximately two dozen employees who were affected by a workforce reduction of 8,000 jobs in May alleges that Meta violated federal antidiscrimination laws. The plaintiffs contend that staff taking protected medical or family leave were unable to maintain continuous AI usage on corporate systems, resulting in depressed metrics that wrongly triggered their termination. Meta has denied all allegations as the legal proceedings continue. Decline of Tokenmaxxing Culture and Internal Metric Competition Prior to the recent guidelines, the institutional emphasis on AI usage created an internal trend among workers seeking to maximize their adoption metrics, a behavior referred to within the company as 'tokenmaxxing'. To elevate their internal standing, some employees continuously submitted repetitive prompts to AI tools solely to boost their recorded consumption of tokens, which represent the baseline unit of AI computational processing. The practice peaked when a Meta employee created an internal tracking dashboard that ranked workers by their AI tool consumption, bestowing labels such as 'Token Legend' upon top consumers. After details regarding the leaderboard circulated externally, the dashboard was deactivated in April. Months later, Meta implemented rationing measures on workforce AI consumption to curb frivolous usage. Testing the Autonomous Hatch AI Agent More recently, Meta has been inviting employees to test its most advanced experimental AI deployment, an agentic system named Hatch. Operating similarly to popular autonomous software projects such as OpenClaw, Hatch possesses the capability to execute multi-step tasks independently, navigate internet browsers, and control third-party computer applications. Workers have been evaluating Hatch on corporate hardware over recent weeks in advance of a potential public deployment. However, some employees remain cautious about connecting Hatch to personal email addresses, digital calendars, and non-work digital accounts. These hesitations stem from privacy concerns as well as risks that the autonomous tool might inadvertently corrupt critical personal data. Employee confidence had previously been damaged by a now-paused Meta initiative that recorded user keystrokes and device activity on corporate computers to harvest AI training data, leaving some workers hesitant about granting broad system access to new tools. Despite these reservations, the assurance that declining to participate in AI testing will not negatively influence performance reviews has bolstered workforce morale. Several employees have begun using Hatch to schedule personal appointments and manage administrative tasks outside of work, describing the software as highly capable. Nevertheless, some workers have raised questions regarding the environmental and financial costs associated with consuming substantial token volumes to automate routine daily activities. Meta declined to comment on the testing of Hatch. Productivity Expectations, Computing Costs, and Layoff Concerns Employees report that despite the end of intentional tokenmaxxing, Hatch consumes significantly more computational tokens than traditional chatbots or automated coding assistants. Workers observe that leadership remains keenly interested in seeing employees demonstrate their ability to effectively leverage AI systems for workplace productivity. If Hatch achieves its intended productivity gains, some employees express concern that management could pursue further workforce reductions. Following the elimination of 8,000 roles in May, reports emerged last week indicating that Meta was preparing for another major layoff round. However, those plans were called off following an increase in software bugs linked to over-reliance on AI tools and slower than expected progress in autonomous agent development. Meta Chief Executive Officer Mark Zuckerberg has publicly stated that additional mass layoffs are not anticipated for the remainder of the current year. What this means for you Meta's decision sets a significant precedent for tech industry workplace policies, employee evaluations, and artificial intelligence integration. • Impact on Tech Workers: Broad tech companies may rethink mandatory AI adoption targets for employee appraisals, relieving staff from forced tool usage. • Workplace Privacy Standards: Privacy hesitations regarding autonomous agents accessing personal accounts highlight the need for stronger corporate data governance. • Software Development Quality: Heightened awareness of software bugs linked to over-reliance on AI will prompt firms to balance automation with human oversight. • Corporate Management Practices: Evaluating workforce productivity through raw consumption metrics like token counts is likely to be abandoned in favor of actual output. Questions & Answers 1. What changes did Meta make to its employee performance review policy? Meta removed explicit references to AI usage and 'AI Native' designations, returning to evaluating employees solely on their actual contributions. 2. What is the new Hatch AI tool being tested by Meta? Hatch is an autonomous AI agent developed by Meta that can browse the web and operate software applications independently on computers. 3. Will Meta continue tracking token consumption for performance ratings? No, Meta leadership confirmed that token counts and AI adoption dashboards will no longer be used to evaluate employee performance. 4. Why were laid-off employees suing Meta over AI usage? A group of employees laid off in May filed a lawsuit alleging that medical and family leave prevented them from accumulating high AI usage metrics, leading to unfair dismissal. 5. Are further mass layoffs expected at Meta this year? Meta CEO Mark Zuckerberg has stated that no additional mass layoffs are anticipated for the remainder of this year. https://trendkia.com/en/ai/meta-ne-karmachariyon-ke-mulyankana-niyamon-men-kiya-badalava-nae-svayatta-hatch-ai-tula-ki-testinga-shuru-26748 TrendKia — Har trend, sabse pehle.