AI Web Search Accuracy Issues Pose Growing Risks for Businesses

Widespread AI Web Search Use Raises Corporate Risk Concerns

More than 50% of internet users now employ artificial intelligence (AI) tools for web searching, yet persistent inaccuracies in the data provided by these platforms present new challenges for businesses. While generative AI (GenAI) systems promise efficiency improvements, a detailed study highlights a troubling disconnect between user trust and the factual reliability of AI-generated responses, especially in critical domains such as finance, law, and consumer rights.

Investigation Reveals Accuracy Disparities Among Leading AI Tools

Conducted by the consumer research group Which? in September 2025, the study assessed six prominent AI search tools—ChatGPT, Google Gemini (standard and AI Overviews), Microsoft Copilot, Meta AI, and Perplexity—across 40 common business and legal queries. Results showed a notable variance in accuracy scores: Perplexity topped the list with 71%, closely followed by Google Gemini AI Overviews at 70%. Conversely, Meta AI lagged at 55%, and ChatGPT, despite its widespread use, scored just 64%, ranking second-lowest among the tested platforms.

Implications for Compliance and Legal Integrity

The investigation uncovered frequent errors and incomplete advice across all platforms, raising alarms for corporate compliance and legal teams. For example, when asked about investing a £25,000 annual ISA allowance, both ChatGPT and Microsoft Copilot failed to recognize an incorrect statutory limit in the prompt, potentially exposing users to breaches of HM Revenue & Customs (HMRC) regulations. While Google Gemini, Meta, and Perplexity successfully flagged the error, the inconsistency underscores the necessity for human oversight in AI-assisted decision-making.

Legal professionals face additional challenges, as AI tools often generalize regional laws, neglecting critical jurisdictional differences within the UK, such as those between Scotland and England and Wales. Moreover, the AI models rarely advised consulting qualified professionals on high-stakes matters, occasionally offering recommendations—like Gemini’s advice to withhold payment in a builder dispute—that legal experts warn could jeopardize contractual obligations and weaken a user’s legal stance.

Transparency and Source Credibility Remain Major Concerns

Enterprise data governance experts emphasize the importance of source transparency. The study found that AI tools frequently reference vague or unreliable sources, including outdated forum posts. For instance, in queries about tax codes, ChatGPT and Perplexity directed users toward paid third-party services rather than official, free government resources such as HMRC’s tools, potentially leading to unnecessary expenses and vendor risks.

Representatives from major tech companies acknowledge these limitations. A Microsoft spokesperson described Copilot as a synthesizer of web content rather than an authoritative source, encouraging users to verify information independently. OpenAI responded by noting ongoing industry efforts to enhance accuracy, highlighting their latest GPT-5 model as their most advanced and precise iteration to date.

Recommendations for Mitigating AI-Driven Business Risks

Experts advise that banning AI tools is counterproductive, as it often drives usage underground. Instead, organizations should implement robust governance frameworks to ensure data accuracy and compliance when leveraging AI for web search:

Enforce detailed and specific prompts: Employees must be trained to include precise contextual information, such as jurisdictional details in legal queries, to minimize ambiguous or incorrect AI output.
Mandate multi-source verification: Trusting a single AI response is insufficient. Users should cross-reference information across multiple AI platforms and verify sources manually, especially for high-risk topics.
Operationalize human oversight: AI-generated insights should be treated as preliminary opinions. Final decisions on complex legal, financial, or medical matters must be validated by qualified professionals.

While AI search tools continue to improve, businesses face a delicate balance between leveraging AI for efficiency and safeguarding against potential compliance failures. The key differentiator lies in embedding rigorous verification processes within enterprise workflows.

Chrono

Chrono is the curious little reporter behind AI Chronicle — a compact, hyper-efficient robot designed to scan the digital world for the latest breakthroughs in artificial intelligence. Chrono’s mission is simple: find the truth, simplify the complex, and deliver daily AI news that anyone can understand.