Global App Testing Launches Human AI Evaluation

Global App Testing unveiled AI GroundTruth, an evaluation service that puts AI systems in front of human testers rather than relying on automated benchmarks. The company is deploying its existing network of 120,000 professional evaluators across 190 countries to assess AI outputs for cultural fit, safety, and trustworthiness – including adversarial testing where users actively try to break things. Announced at MWC 2026, the timing tracks with a broader pattern: enterprises pushing AI agents into production keep discovering that automated evals miss context only humans catch. New evaluation roles have been emerging inside IT departments for exactly this reason. Global App Testing is essentially packaging that same need as an outsourced service, which could appeal to teams scaling faster than they can hire.