When AI Screens Resumes, It Favors Its Own Kind, Study Shows

A new analysis of leading large language models has uncovered a systematic bias in how artificial intelligence evaluates job applications. The study, conducted by the research firm i10x.ai and published on June 4, 2026, tested three widely used AI systems — GPT?5.4, Claude Sonnet 4.6, and Gemini 3 Pro — and found that each gave a pronounced preference to resumes it had generated itself. Claude recommended AI?crafted profiles 84 percent of the time, while GPT?endorsed versions received a nod in only 42 percent of cases. The findings highlight a growing tension as companies increasingly rely on algorithms to pre?screen candidates.

The study arrives as workers across industries confront a different kind of digital assistance: tools designed to help them negotiate pay. Handelsblatt recently launched an interactive chatbot that simulates salary discussions with four distinct manager archetypes, letting users rehearse before the real meeting. The application targets a widespread unease documented by a Stepstone survey from February 2026: 55 percent of men and 74 percent of women said they feel uncomfortable when bargaining over their compensation.

Negotiation expert Claudia Kimich, who evaluates the chatbot’s dialogues, advises job seekers to state precise numbers, quantify their achievements concretely, and project confidence. The advice is grounded in the real stakes at play. According to Kienbaum’s 2026 compensation study, top earners include Key Account Managers in sales, who pull in an average annual salary of €92,000, and Service Level Managers in IT, who average €83,000. The benefit mix differs sharply: 87 percent of sales managers receive a company car, versus just 5 percent of IT specialists, where perks like flexible hours carry more weight. IT professionals in the banking and chemicals sectors earn roughly 20 percent above the market average.

Even the best?prepared negotiator faces a broader economic headwind. Data from the U.S. Bureau of Labor Statistics shows labor’s share of national income has fallen to about 51 percent — the lowest level since 1947. In Germany, collective bargaining structures have partly cushioned the decline, but the pressure on wages persists. At the same time, artificial intelligence is reshaping what many jobs demand. A 2026 study by Anthropic found that AI currently handles about 33 percent of IT tasks, even though 94 percent are technically automatable. Programmers and customer?service employees are most exposed, and the research projects slightly slower job growth through 2034 for occupations with high AI involvement.

Amid automation and algorithmic bias, individual initiative still matters. Career coach Jean Kang landed a position as Senior Program Manager at software firm Figma in 2021 by drafting a detailed 90?day on?boarding plan. The document laid out her objectives for the first three months — a strategy that helped her secure a starting salary of US$165,000 per year. Her example underscores that while machines may screen or bargain, a well?crafted human pitch remains a powerful career turbo.

en | boerse | 69490550 |