Okta CEO Todd McKinnon is warning about the security problems with AI

Okta CEO Todd McKinnon is warning about the security problems with artificial intelligence as it shifts from test models to being used in real situations. He strongly emphasized the urgent need for identity-based security systems that can control what AI agents can access.

In a recent interview with CNBC’s Jim Cramer, McKinnon emphasized that as AI agents transition from prototypes to production environments, they introduce novel cybersecurity challenges that organizations must urgently address.

AI agents now work in real systems and must have safe, controlled access

McKinnon told companies using AI agents in real business settings to treat the systems like coworkers. He explains that moving AI into “production” gives these agents real access to live systems that hold private customer information, internal files, and business data. 

This comes with big security risks that companies can’t afford to ignore.

McKinnon warns that AI systems can now act independently and make unsupervised decisions that could lead to unauthorized access, data leaks, or accidental misuse of information. He advises that the only way to handle these risks safely is by using monitoring systems that check who or what is trying to access data. These systems only allow access if that individual or AI agent is trusted, verified, and has permission.

McKinnon says that Okta’s identity management lets businesses manage, track, and control access for all types of users, including AI agents. 

Okta remains careful with its forecasts even after reporting strong earnings

Okta reported better-than-expected earnings and revenue for the quarter but didn’t raise its revenue forecast for upcoming quarters. Instead, Okta maintained a cautious and conservative outlook that dropped the company’s stock by more than 11% during after-hours trading.

McKinnon said it is difficult to plan future growth with full confidence because of the unpredictable global economic climate. 

He said Okta’s leadership will stay grounded, plan carefully, and protect the company’s position rather than chase short-term goals too aggressively. 

This way, Okta will adjust when market conditions in industries where customer budgets and tech investment cycles keep shifting change.

McKinnon also pointed out that investors may feel frustrated when companies like Okta don’t immediately speed up growth after reporting strong results. 

Todd said companies must avoid mistakes when building secure, scalable identity systems that work across human users and AI systems. 

Artificial intelligence companies have been urged to replicate the safety calculations underpinning Robert Oppenheimer’s first nuclear test before releasing all-powerful systems.

Max Tegmark, a leading voice in AI safety, said he had carried out calculations akin to the US physicist Arthur Compton before the Trinity test and had found a 90% probability that a highly advanced AI would pose an existential threat.

McKinnon also explained that forcing growth too quickly will make it difficult for a company to support new customers properly. Still, he noted his excitement about where the company is headed and emphasized that patience and discipline keep a company strong. Okta is thinking beyond the next quarter and is preparing to lead in a future where AI and secure identity management will go hand in hand.

Your crypto news deserves attention - KEY Difference Wire puts you on 250+ top sites