A new study conducted by US-based digital marketing agency Legal Guardian Digital has reignited debate over which artificial intelligence chatbot users can trust the most. While OpenAI’s ChatGPT remains the world’s most widely recognised AI assistant, the latest rankings suggest that popularity may no longer guarantee reliability.
According to the report, Perplexity AI emerged as the most reliable AI chatbot among major global competitors, outperforming ChatGPT, Google’s Google Gemini, Anthropic’s Claude, and several other AI systems.
The findings come at a time when businesses, students, researchers, and everyday users are increasingly relying on AI assistants for information, productivity, coding, customer support, and decision-making. As AI adoption accelerates globally, concerns surrounding hallucinations, factual inaccuracies, downtime, and user trust have become central to evaluating chatbot performance.
The report assessed multiple AI assistants using several practical benchmarks designed to measure real-world reliability. Researchers examined:
The goal was to determine which chatbot was least likely to provide fabricated or misleading information while remaining consistently available for users.
AI hallucinations — instances where chatbots confidently generate false or inaccurate information — have become one of the industry’s biggest concerns. As AI tools become integrated into workplaces, education systems, healthcare, and financial services, accuracy is increasingly viewed as more important than raw popularity.
According to the US-based digital marketing agency Legal Guardian Digital's study, Perplexity AI achieved the strongest overall reliability performance with a score of 85 out of 100.
The platform reportedly recorded:
This was the lowest among all major AI chatbots included in the study and significantly below the industry average of 22 per cent.
Researchers stated that users experienced no service outages during the testing period, helping Perplexity gain a major advantage over competitors.
Perplexity has steadily gained attention in recent years for combining conversational AI with live web-based search functionality. Many users prefer the platform because it often provides citations and source references alongside AI-generated responses.
Industry experts believe this search-integrated model may help reduce hallucination risks compared to traditional standalone large language models.
xAI’s Grok ranked second in the reliability rankings.
The report stated that Grok achieved:
The strong performance highlights the growing competitiveness of newer AI companies entering the market.
Chinese AI assistant DeepSeek ranked third overall with:
Researchers also highlighted that DeepSeek remains free to use despite outperforming several premium AI platforms.
The rise of DeepSeek reflects the rapidly intensifying global AI race, particularly between American and Chinese AI developers.
Despite being the most widely used AI chatbot globally, ChatGPT ranked sixth in the reliability index.
Researchers claimed that the chatbot generated inaccurate responses in nearly 30 per cent of cases — almost double the error rate reported for DeepSeek.
However, ChatGPT continued to perform strongly in user satisfaction metrics.
The study found that ChatGPT maintained a user satisfaction rating of 4.7 out of 5, demonstrating that users still value its conversational fluency, creativity, and versatility.
OpenAI’s chatbot continues to dominate in areas such as:
Still, the report suggests that users are increasingly prioritising factual reliability alongside functionality.
Anthropic’s Claude reportedly ranked seventh in the study and experienced more outages than several competitors.
Claude has often been praised for safety-focused AI design and long-form reasoning capabilities. However, the report indicates that uptime reliability may still remain a challenge.
Google Gemini secured eighth place, while Meta’s Meta AI ranked ninth overall.
The findings may increase competitive pressure on major technology firms as the AI industry matures and users become more selective.
The study arrives during a period of growing global concern around AI misinformation, fabricated content, and unreliable chatbot outputs.
As AI assistants become integrated into everyday workflows, industries are demanding stronger safeguards against hallucinations and incorrect recommendations.
According to market intelligence firm Sensor Tower, US uninstalls of ChatGPT’s mobile app surged by 295 per cent on February 28 compared with the previous day. The report noted that ChatGPT’s typical daily uninstall rate had averaged around 9 per cent over the previous month.
These figures suggest that some users may be experiencing growing frustration or fatigue with AI-generated inaccuracies.
Austin Hunt, chief executive of Legal Guardian Digital, said many users still assume ChatGPT is automatically the best AI assistant because it entered the market early and became the most famous.
However, the latest rankings indicate that newer AI platforms are rapidly improving in areas that matter most to users, including reliability, stability, and factual accuracy.
As AI adoption continues expanding across enterprises, governments, education systems, and financial services, industry analysts believe trustworthiness could become the defining factor separating market leaders from competitors.
The latest AI reliability study highlights a major shift in the competitive landscape of artificial intelligence chatbots. While ChatGPT remains the most recognised and widely used AI assistant globally, newer platforms like Perplexity AI, Grok, and DeepSeek are gaining momentum by focusing heavily on accuracy, uptime stability, and reduced hallucination rates.
Perplexity’s top ranking demonstrates that users increasingly value reliable and fact-based responses rather than brand recognition alone. As AI systems continue becoming embedded into critical workflows and decision-making environments, trust and consistency are likely to emerge as the most important competitive advantages.
The findings also underline how rapidly the AI industry is evolving. With companies racing to improve accuracy, safety, and real-time performance, the next phase of the AI battle may be defined less by hype and more by dependability.