{"id":87101,"date":"2025-05-12T13:00:00","date_gmt":"2025-05-12T20:00:00","guid":{"rendered":"https:\/\/www.salesforce.com\/?p=87101"},"modified":"2025-08-18T10:53:14","modified_gmt":"2025-08-18T17:53:14","slug":"combating-ai-hallucinations","status":"publish","type":"post","link":"https:\/\/www.salesforce.com\/news\/stories\/combating-ai-hallucinations\/","title":{"rendered":"Less Hallucinations, More Trust: Salesforce\u2019s Path to Consistent Enterprise-Ready AI"},"content":{"rendered":"\n

We\u2019re living in an era of extraordinary AI advancement. Generative models have evolved from novelties to necessities almost overnight \u2014 capable of drafting complex briefs, summarizing nuanced customer interactions, and reasoning across text, images, and code. But while large language models (LLMs)<\/a> can sometimes deliver high-quality outputs, these are not always optimized for the best business outcomes, and worse, they don\u2019t do so consistently. That inconsistency makes them unreliable to business leaders looking to deploy AI agents<\/a> within the enterprise. <\/p>\n\n\n\n

At the heart of the issue is hallucination: when a model confidently generates false, misleading, or entirely fabricated information. In consumer settings, these mistakes can be meme-worthy. In the enterprise, they raise red flags. According to The New York Times<\/em><\/a>, newer AI systems hallucinate more<\/em>, not less than older models, at rates as high as 79% in some tests. That\u2019s not just ironic \u2014 it points to a deeper flaw in how we measure and understand intelligence in AI<\/a>.<\/p>\n\n\n\n

At Salesforce, we\u2019ve studied this challenge through the lens of real-world business use and identified a pattern that the industry refers to as jagged intelligence<\/a>. It describes the uneven capabilities of today\u2019s models \u2014 systems that can excel at complex, high-level tasks, yet stumble on more basic ones. It\u2019s like hiring an intern who writes brilliant code but forgets to save the file. An LLM can accurately summarize a multi-threaded support case, but follow up with an irrelevant product recommendation. It can cite policy documents chapter and verse, but misses that it\u2019s referencing outdated guidance.<\/p>\n\n\n\n

This unevenness creates a core tension \u2014 AI that seems smart enough to believe, but not consistent enough to fully trust. <\/p>\n\n\n\n

Tensions around hallucinations stand in sharp contrast to the ambiguous promise of artificial general intelligence (AGI) that so often dominates \u200cindustry conversations. That\u2019s why we believe it won\u2019t be AGI, but Enterprise General Intelligence (EGI)<\/a> that represents a tangible North Star: AI purpose-built for business, designed not only for advanced capability, but for consistency \u2014 delivering reliable performance, seamless integration, and trustworthy outcomes across complex systems. <\/p>\n\n\n\n