Are AI tools mostly replacing workers rather than helping them in female-dominated jobs, and is that about the work itself?

This explores whether AI is being used to replace workers rather than assist them in female-dominated jobs, and what drives that split. The corpus has no note that tests this directly. None of the retrieved material compares automation with augmentation by occupation gender, so it can't confirm the premise. It does have some pieces of the surrounding picture.

The closest evidence suggests the driver is the kind of work, not who does it. Workers have handed tasks to AI in structured workflows mostly in information-intensive occupations, and this tracks what the technology can do rather than what people expected to be routine and automatable (Where have workers actually delegated tasks to AI?). If that's right, a gender pattern would come from which occupations are heavy on information tasks, and the corpus can't say whether that's the mechanism here. Firms also differ in how fast they swap people for AI. Firms with more AI exposure replaced online-marketplace workers faster and more cheaply than less-exposed firms (Do firms substitute labor for AI at different rates?). That points to substitution being a business decision that gets easier with scale, not something built into the technology.

The line between automation and augmentation is also blurrier than the question assumes. Research on 8,356 workplace AI risk scenarios found that augmentation isn't inherently safe, because leaning on AI agents can gradually wear down workers' skills and their ability to oversee the AI (Does AI augmentation protect workers from skill erosion?). So a job that is nominally augmented can still slide toward replacement. That fits the broader argument that incremental AI adoption gradually removes society's dependence on human labor (Does incremental AI replacement erode human influence over society?).

The only gender-specific finding is a different kind of evidence. GPT-3.5 refused requests at different rates for female, younger, and Asian-American personas (Do AI guardrails refuse differently based on who is asking?). That shows AI tools can treat users differently by demographics, but it says nothing about whether employers automate or augment. To answer the question you'd need labor-market studies that split AI use by occupation gender and by task type, and this set of notes doesn't include one.