Artificial intelligence is beginning to expose a weakness in one of the fundamental assumptions of performance management: that performance belongs to the individual. Traditional pay for performance models assume that employees create value primarily through their own effort, judgement and expertise. AI challenges that assumption. Increasingly, work is produced through a combination of human capability and machine intelligence, making it difficult to distinguish between the contribution of the employee and that of the technology.

A recent article in The Financial Times raises an important question for employers. If AI enables employees to perform at a higher level, what exactly is being measured? Is the organisation rewarding the employee’s capability, the capability of the AI, or the combination of the two? The same question applies inside organisations. An employee who uses AI to produce higher quality analysis, automate routine work or make better decisions may appear to outperform colleagues who have similar experience but make less effective use of AI. Traditional performance measures capture the outcome but rarely distinguish how that outcome was achieved.

The challenge becomes greater as employees move beyond using AI as an assistant and begin orchestrating multiple AI agents to complete increasingly complex work. Their role shifts from performing tasks themselves to directing, validating and governing a digital workforce. Success depends less on personal output and more on selecting the right AI tools, designing effective workflows, applying human judgement, managing risk and deciding when human intervention is required. The employee is no longer simply completing work. They are creating a system capable of producing work at scale.

This suggests that organisations may need to rethink how they define performance. Rather than measuring only individual output, future performance management may need to distinguish between four complementary dimensions. The first is human capability: judgement, creativity, ethical decision making, leadership and collaboration. The second is AI capability: the effectiveness of the digital tools and agents operating under the employee’s direction. The third is human AI orchestration: how successfully employees combine technology with human expertise to improve quality, speed and decision making. The fourth is capability creation: the lasting organisational value created through reusable workflows, prompts, knowledge assets and AI enabled processes that continue generating value beyond the individual’s own work.

The Financial Times article highlights another emerging dimension that organisations have yet to address. Employees are increasingly using AI to amplify their own knowledge, while employers are asking employees to codify that same knowledge into AI systems through prompts, workflows and documented expertise. In effect, employees are both consumers and creators of organisational intelligence. Current reward systems largely ignore this distinction, treating the development of AI assets as simply part of the job. Yet these assets may continue creating value long after the employee has left the organisation. This raises important questions about intellectual contribution, ownership and whether capability creation should itself become part of performance evaluation.

The implication is that AI does not simply make employees more productive. It changes the nature of performance itself. Organisations are no longer evaluating individuals working alone, but human AI systems that combine expertise, technology and judgement. As AI becomes embedded in everyday work, the challenge for reward professionals will not be measuring performance alone but determining whose performance they are rewarding. The future of pay for performance may therefore depend less on measuring what employees produce and more on understanding how effectively they create value through the partnership between human capability and artificial intelligence.