OpenAI CEO Urges U.S. Action on AI Superintelligence

TLDR

• OpenAI Chief Executive Sam Altman urged U.S. policymakers to prepare for advanced AI systems that are rapidly entering daily economic use.
• He said current AI models already perform coding and research tasks once handled by teams of programmers.
• Altman warned that more powerful systems could enable major scientific discoveries while also increasing cyber and biosecurity risks.
• He cautioned that a “world-shaking cyberattack” could occur as early as this year without strong defenses.
• Ledger CTO Charles Guillemet said AI tools now help attackers find and exploit software flaws within seconds.

OpenAI Chief Executive Sam Altman urged U.S. lawmakers to prepare for advanced artificial intelligence risks and gains. He said current systems already perform coding and research tasks once handled by teams. He warned that policymakers must act quickly as AI moves into daily economic use.

OpenAI Calls for Early Policy Action on Superintelligence

Altman told Axios that AI systems now complete technical work once assigned to groups of programmers. He said newer models will help scientists make discoveries and let individuals perform team-level tasks. He added that policymakers must respond before more advanced systems emerge.

He said the United States must build “superintelligence” aligned with democratic values before rivals succeed. He argued that nationalization would likely fail as a government project. He stated, “That probably wouldn’t work as a government project, I think that’s a sad thing.”

Altman said companies building advanced systems must work closely with the U.S. government. He described OpenAI’s proposals as a starting point for public debate. He stressed that high integrity and trustworthy leadership remain essential in AI development.

He also said AI will function like electricity and embed across devices. He explained that basic intelligence costs will fall while top systems remain expensive. He said users may pay higher monthly bills when they use more advanced services.

Cybersecurity and Crypto Face Rising AI-driven Threats

Altman said AI could enable stronger cyberattacks and lower barriers to harmful biological research. He warned that a “world-shaking cyberattack” could occur as early as this year. He said preventing such an event would require a “tremendous amount of work.”

He also warned that open-source models may soon handle complex biology tasks. He said society must prepare for groups that attempt to create novel pathogens. He stated, “The need for society to be resilient… is no longer a theoretical thing.”

Charles Guillemet, chief technology officer at Ledger, said AI tools now reduce the cost of exploiting software flaws. He told CoinDesk that attackers can reverse-engineer code and link vulnerabilities within seconds. He said prompts now replace months of manual work.

The crypto sector lost over $1.4 billion in hacks and exploits last year. Guillemet said growing reliance on AI-generated code may introduce flaws at scale. He urged stronger defenses, including mathematically verified code and offline hardware wallets.

Altman also pointed to shifts in labor patterns during 2026. He said programmers already work differently compared with last year. He predicted continued changes as AI tools expand across industries.

He added that greater processing power may lower costs as demand rises. He said energy capacity will support expanded AI infrastructure. He maintained that coordination between companies and government remains critical as systems advance.

The post OpenAI CEO Urges U.S. Action on AI Superintelligence appeared first on Blockonomi.