In a CX Today interview, industry expert Fornes contrasts traditional chatbots with agentic AI, likening the latter to a “wise owl” capable of executing complex actions, such as accessing bank accounts and managing sensitive data. He warns that consumers and enterprises are acting as “unpaid QA analysts,” unwittingly sharing credentials with systems they don’t fully comprehend. Fornes likens this to handing a wallet to a stranger claiming to offer cheap flights, highlighting a legal void where liability for AI errors currently rests on the user, not the software provider. Furthermore, he discusses the unique challenges in securing these AI systems, which rely on linguistic interpretations, making them prone to prompt injections. Fornes advises business leaders to treat agentic AI like “a talented but reckless genius,” signifying its potential while emphasizing the need for cautious handling of critical company data. This highlights the importance of trust, security, and understanding in AI adoption.
Source link
