I am a CEO of cybersecurity which advises more than 9,000 agencies and Sam Altman is wrong that the crisis in Fraud in AI is coming – that’s already there

This already happens and it does not only come for banks; He is hitting each part of our government right now.

Each week, the fraud generated by the siphon of millions of public services systems, rescue funds in the event of disaster and unemployment programs. Criminal networks already use Deepfakes, synthetic identities and large languages models to go beyond the defenses of overwhelmed fraud, including easily usurped tools, such as facial recognition, and they win.

We saw an overview of this during the pandemic, when the fraud rings exploited gaps in state systems to steal hundreds of billions of unemployment benefits. It was not only people who wore masks to bypass facial recognition. These were false identities generated by AI, vocal clones and forged documents of overwhelming systems that were not built to detect them. Today, these tactics are more advanced and fully automated.

I work with more than 9,000 agencies across the country. As I testified to the House of American Representatives twice this year, what we see on the field is clear. The fraud is faster, cheaper and more scalable than ever. The groups of organized crimes, both national and transnational, use a generative AI to imitate identities, generate synthetic documentation and flood our fraudulent complaint systems. They not only steal the government; They fly to the American people.

The Small Business Administration Inspector General now estimates that nearly $ 200 billion has been stolen from the chemage insurance programs in the pandemic era, making it one of the biggest fraud losses in the history of the United States. Medicaid, IRS, Tanf, Chip and Catastrophe programs are faced with similar vulnerabilities. We have also seen this first in our work alongside the American secret services protecting the Snap USDA program, which has become a buffet for fraudsters with billions stolen nationally each month. In fact, in a single day using an AI, a fraud ring can deposit tens of thousands of false complaints in several states, most of which will be treated automatically unless you are reported.

We reached a turning point. While AI continues to evolve, the scale and sophistication of these attacks will increase quickly. Just as Moore’s law predicted that computer power would double every two years, we now live through a new type of exponential growth. Gordon Moore, co-founder of Intel, initially described the trend in 1965, and it guided decades of innovation. I believe that we can soon recognize a similar principle for AI that I call the “Altman law”: every 180 days, the capacities of AI double.

If we did not modernize our defenses with the same pace as technological progress, we will be surprised permanently.

What we desperately need are smarter tools and infrastructure, no more bureaucracy.

This means superimposing the advanced verification of the identity, not only analyzes or passwords of the face. This means using real-time data, behavioral analyzes and trans-jurisdictional tools that can point out anomalies before money leaves the door. This also means reliving what has already worked: tools such as national clarification clearinghouse, which reported billions of dollars of double services through the state lines before its closure.

AI is a multiplier of strength, but it can be more easily armed than it can be exercised for protection. Criminals are currently using it better than us. Until it changes, our most vulnerable systems and people who depend on them will remain exposed.

The opinions expressed in the Fortune.com comments are only the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.