Most businesses are now fully embracing AI – but aren’t always protected against the risks

Gallagher survey finds high adoption of AI in workplaces
However many firms still have major concerns about the technology
Hallucinations and legal liability top the list of worries

Businesses are adopting AI tools at a faster rate than ever before, but many still have some significant concerns about its effects, new research has claimed.

The 2025 Attitudes to AI Adoption & Risk Survey from Gallagher found almost half (45%) of businesses were using AI in day-to-day operations by the end of 2024, an increase of a third (32%) over the previous 12 months.

It added 85% of businesses have introduced strategies related to job protection as part of their AI adoption framework, as seven of the top 10 challenges associated with AI adoption relate to how to change business strategy to effective embrace AI.

AI concerns

Gallagher’s research found larger firms were particularly keen at getting AI up to speed, with 82% reporting success in adopting AI, with improved efficiency and productivity named as particular benefits.

The top use-cases for AI named by survey respondents included writing emails/agendas(32%), handling customer queries (31%), and understanding market dynamics (28%).

However the survey also highlighted some widespread concerns about the use of AI, particularly as Gallagher predicts a possible increase in legal liabilities for firms where the use of AI is not properly governed over the next few years.

AI errors or hallucinations, where a model generates inaccurate results, was the most common threat, reported by more than a third (34%) of business owners, with data protection and privacy violations (33%) and legal liabilities (31%) related to AI misuse also mentioned.

Gallagher has warned companies utilizing AI to be aware of the possible risks using AI systems could expose them to, and take the necessary precautions where needed

“AI systems excel at processing vast amounts of data and identifying patterns that may not be readily apparent to humans,” noted Ben Waterton, Executive Director of Professional Indemnity at Gallagher.

“However, they cannot replace human expertise and judgment that qualified individuals bring to their work. Relying solely on AI without critical examination and human oversight can lead to serious consequences and compromising advice. The quality assurance procedures and oversight of employees must evolve to ensure that this emerging risk is recognised and addressed to prevent professional indemnity losses.”