Altman Warns of Impending Fraud Crisis Driven by AI Exploitation

OpenAI’s CEO, Sam Altman, has warned that the world may be on the brink of a «fraud crisis» as artificial intelligence enables malicious actors to impersonate others more convincingly.

«I am alarmed that there are still financial institutions that accept voiceprints as a means of verification for large transfers or other operations. You just say a passphrase, and everything happens automatically,» Altman remarked. «This is madness… AI has entirely displaced most forms of authentication aside from passwords.»

These statements were made during his extensive interview discussing the impact of AI on the economy and society with the Federal Reserve on Tuesday. He also addressed an audience of representatives from major U.S. financial institutions about AI’s role in the economy.

His comments come ahead of the anticipated release of the White House’s «AI Action Plan» in the coming days—a roadmap addressing the regulation of this technology and the goal of maintaining America’s leadership in the AI sector.

Despite Altman’s concerns regarding the risks associated with this technology, OpenAI recently urged the Trump administration to reject regulations they believe could inhibit tech companies from competing with foreign AI developers. Earlier this month, the U.S. Senate voted to remove a contentious provision from Trump’s bill that would have prohibited states from enacting AI-related laws for a decade.

Altman is not alone in his worries over how AI could contribute to rising fraud.

Last year, the FBI warned about scams involving AI-generated voice and video. Many parents reported that AI voice technologies were used to deceive them into believing their children were in trouble and extort money from them. Moreover, U.S. officials recently alerted the public that someone was using AI to mimic Secretary of State Marco Rubio’s voice, contacting state department ministers, governors, and members of Congress.

«I am very concerned that we are on the verge of a serious fraud crisis,» Altman said. «Right now, it’s a voice call, but soon it’ll be a video call or FaceTime that you won’t be able to distinguish from reality,» he added.

Altman also expressed his fears regarding the prospect of malicious actors creating an «AI superintelligence» and using it irresponsibly before the rest of the world becomes advanced enough to defend against such an attack.

Additionally, he voiced concerns about the potential loss of control over superintelligent AI systems or granting excessive decision-making power to the technology. Various tech companies, including OpenAI, are working to develop such superintelligent AI, and Altman believes that by the 2030s, AI will surpass human capabilities. However, it remains unclear how they define this milestone or when it may be reached.

In contrast to some of his Silicon Valley peers, Altman expressed less concern over AI’s impact on the job market. CEOs like Anthropic’s Dario Amodei and Amazon’s Andy Jassy have warned that this technology could lead to widespread job losses.

«There are many seemingly sensible predictions—‘Oh, this will happen to this and that, and this will affect the economy.’ Nobody knows. To me, it’s too complex a system, and it’s too new and influential a technology to predict,» he stated.

He does have an opinion on the matter. Altman mentioned that while «many professions may disappear,» new ones will emerge. He reiterated his earlier prediction that if we could look 100 years into the future, future workers might not have what we consider «real jobs» today.

Simultaneously with Altman’s remarks, OpenAI released a report from its chief economist, Roni Chatterji, which outlines the productivity benefits of ChatGPT.

Chatterji, who joined OpenAI after serving as the CHIPS and Science Act coordinator in the Biden administration, compared AI to revolutionary technologies like electricity and transistors. He noted that ChatGPT already has 500 million users worldwide.

According to the report, 20% of users in the U.S. employ ChatGPT as a «personal tutor» for «learning and upskilling,» although the report does not specify what exactly users are learning through the service. Chatterji also pointed out that more than half of ChatGPT users in the U.S. are aged 18 to 34, indicating that the use of AI tools in the workplace could yield long-term economic benefits.

*Would you like to stay updated on important AI news? Subscribe to our Telegram channel* [BotHub AI News](https://t.me/bothub).

For the original news source, click [here](https://edition.cnn.com/2025/07/22/tech/openai-sam-altman-fraud-crisis).