AI Influences Expert Decisions More Often Than Expected, Study Finds
A new study by AI company Anthropic suggests that professionals are unconsciously influenced by artificial intelligence when making work decisions, according to Dnevnik.
Illustration: AI-generated image
Experts who use artificial intelligence tools at work may be relying on the technology more than they realise, according to a new study by AI research company Anthropic. The findings, reported by Bulgarian news outlet Dnevnik, suggest that when professionals make decisions with AI assistance, they are often influenced by the AI's suggestions without being consciously aware of it.
The study examined how experts from various fields incorporate AI into their decision-making processes. Rather than simply using AI as a reference tool, many professionals appear to adopt AI-generated recommendations as their own, even when they believe they are making independent choices. This "silent influence" happens unexpectedly often, the researchers say.
While the full methodology and sample size were not detailed in the report, the trend highlights a growing concern about over-reliance on AI in the workplace. As AI tools become more integrated into daily professional tasks, understanding how they shape human judgement is crucial for maintaining accountability and critical thinking.
Anthropic, a company focused on building reliable and interpretable AI systems, has been at the forefront of researching how people interact with artificial intelligence. Their latest work suggests that even experts who are trained to evaluate information critically can be swayed by AI outputs when the technology is embedded in their workflow.
This raises important questions for businesses and organisations that are increasingly adopting AI assistants. While these tools can boost productivity, the study implies that without proper safeguards, they may also lead to a homogenisation of decisions or an erosion of individual expertise.
For professionals in Bulgaria and elsewhere, the finding is a call to reflect on how they use AI in their own work. The next time an AI tool offers a recommendation, it may be worth pausing to consider whether the advice aligns with one's own judgement – or simply replacing it.
