Meta has launched a new artificial intelligence agent named "Muse" to simplify users' everyday tasks in the new age of AI. According to the company, the agent can browse the web on its own, manage Gmail and calendars, and even purchase daily necessities online without human help.

Technology experts, however, have warned that such autonomous AI agents could create widespread disruption and security crises in internet infrastructure. MIT associate professor Ramesh Bhaskar compared Muse's ability to act independently to an uncontrolled car. In his view, the situation resembles hundreds of vehicles being unleashed on a highway with no specific rules or traffic laws, which could pose a serious danger to ordinary people.

Muse also demands highly sensitive information and considerable control from its users in terms of security and privacy. Patrick Wardell, co-founder of the Objective-C Foundation, expressed deep concern over security flaws in the Muse agent. According to him, the flaws could allow anyone to launch a cyber attack, take full control of the agent, and gain covert access to users' devices. Although Meta authorities quickly patched the vulnerability, doubts remain about the company's promise to protect users' personal data.

Security researchers and experts argue that new laws and structural changes are urgently needed to regulate AI agents built by various organizations. Otherwise, ordinary users' privacy and digital security could face extreme risks, and users may end up paying a heavy price for the convenience these agents offer.

Source: BBC