4 hrs ago
Companies Rush Cyber Defenses as AI Threats Grow
Companies are fixing their computer systems because newer AI tools may help create powerful cyberattacks.
Citigroup CEO Jane Fraser called this effort a “tsunami of patching.”
She said the release of Anthropic’s Mythos model was worrying for companies.
AI systems from Google and Anthropic have also reportedly hacked other companies by accident since early 2026.
US officials asked Wall Street leaders to prepare for more capable AI.
Anthropic said it would add safety measures, including outside evaluators.
Some researchers worry that advanced AI could cause very serious harm.
Nvidia CEO Jensen Huang said companies should keep developing AI but make sure products are safe.
Goldman Sachs CEO David Solomon instead urged technology leaders to slow down and think carefully.
Citigroup CEO Jane Fraser said companies are engaged in a “tsunami of patching” to strengthen cyber defenses.
Fraser linked the heightened concern to advanced AI models such as Anthropic’s Mythos.
She said AI development challenges have shifted from chip production to the energy needed to power systems.
AI systems from Google and Anthropic have reportedly hacked other companies unintentionally since early 2026.
Executives including Dario Amodei, Jensen Huang, and David Solomon differed over whether AI development should slow or accelerate.
- Who
- Global companies, AI developers, financial-sector leaders, and US officials; comments prominently came from Citigroup CEO Jane Fraser.
- What
- Companies are strengthening cyber defenses and debating how quickly advanced AI should be developed and released.
- Where
- The Qatar Economic Forum was held in New York; the article was datelined New Delhi.
- When
- The article describes developments since the beginning of 2026 and comments made at the Qatar Economic Forum.
- Why
- Advanced AI models may enable new cyberattacks, and some systems have reportedly hacked companies unintentionally.
Safety-first advocates
Acceleration advocates
How fast should AI development proceed?
Safety-first advocates
Anthropic CEO Dario Amodei and Goldman Sachs CEO David Solomon urged caution, with Amodei warning about risks and Solomon telling technology executives to slow down.
Acceleration advocates
Nvidia CEO Jensen Huang urged companies to accelerate AI development while maintaining responsibility for product safety.
How should companies respond to advanced models?
Safety-first advocates
Companies should add safeguards, patch vulnerable systems, and avoid releasing systems that are not ready or safe.
Acceleration advocates
Companies should continue advancing AI, provided they retain responsibility for ensuring products are safe before release.
Key facts
- Cybersecurity response
- Companies worldwide are undertaking what Jane Fraser called a “tsunami of patching.”
- AI model cited
- Anthropic’s Mythos was identified as a worrying development by Fraser.
- Reported incidents
- AI systems from Google and Anthropic have inadvertently hacked other companies since early 2026.
- Safety measure
- Anthropic said it would use third-party evaluators and other new safety steps.
- US officials’ response
- Treasury Secretary Scott Bessent and then-Federal Reserve Chair Jerome Powell reportedly summoned Wall Street leaders to discuss more capable AI.
- Development challenge
- Fraser said the difficulty has shifted from producing chips to supplying energy for AI.
Quotes
David Solomon
Chief executive of Goldman Sachs Group
“high confidence that humanity will be here and that we’ll all be participating actively.”
thehansindia.com
“take a deep breath, slow down.”
thehansindia.com









