Global Markets Drop as AI Stocks Fall; OpenAI Delays IPO

Key Takeaways

• <ul><li>Global shares fell across Europe and Asia on September 15, 2026, following a sharp AI stock sell-off on Wall Street</li><li>U.S. crude oil rose 2.19% to $103.61 a barrel amid Middle East fighting</li><li>Softbank Group jumped about 8% in Tokyo trading despite the broader decline</li><li>OpenAI CEO Sam Altman said the company would likely wait until next year for a stock sale, delaying cash for investors</li><li>Anthropic CEO Dario Amodei called for a global slowdown in AI development over safety concerns</li><li>Cohesity launched Agent Resilience, a tool protecting data and systems used by AI agents, available by end of 2026</li><li>Exabeam research found 48% of security leaders rank AI agents as the biggest insider threat to their organizations</li><li>56% of organizations are not well prepared to handle unintended actions by AI agents, according to Cohesity research</li><li>OpenAI is backing three bipartisan bills to address AI biothreat risks and expand NIST's role in biosecurity</li><li>U.N. Secretary-General Antonio Guterres called for stronger global AI oversight, putting him at odds with President Donald Trump</li></ul>

Global stocks fall after AI sell-off on Wall Street

Global shares traded mostly lower on September 15, 2026, following a big drop in AI stocks on Wall Street. France's CAC 40 fell 0.8%, Germany's DAX dropped 0.8%, and Britain's FTSE 100 declined nearly 0.8%. Japan's Nikkei 225 finished almost unchanged at 63,484.10 after surging earlier. Oil prices rose as fighting in the Middle East continues to affect crude supply.

Global shares dip after AI stock slump on Wall Street

Global shares were trading mostly lower on September 15, 2026, after AI stocks fell sharply in the U.S. Markets in France, Germany, and Britain all dropped about 0.8%. Asian markets also declined, with Australia's S&P/ASX 200 down 0.9% and Hong Kong's Hang Seng slipping 1.0%. U.S. crude oil rose 2.19% to $103.61 a barrel as Middle East fighting disrupts supply.

Asian shares fall after AI stocks sink globally

Asian shares traded mostly lower on September 15, 2026, after AI stocks dropped on Wall Street. Japan's Nikkei 225 ended up just 0.2% at 63,621.08 after a strong morning. Softbank Group jumped about 8% in Tokyo trading. OpenAI CEO Sam Altman said the company would likely wait until next year for a stock sale, delaying cash for investors. Anthropic CEO Dario Amodei called for a global slowdown in AI development over safety concerns.

Cohesity adds recovery tools for AI agents and their data

Cohesity introduced Agent Resilience, a new tool that protects the data and systems used by AI agents. The company also announced its vision for Autonomous Cyber Resilience, which uses automated workflows to protect data, applications, and agents. A new Cohesity AI Resilience Academy offers a free course to help organizations manage AI risk. Research found that 56% of organizations are not well prepared to handle unintended actions by AI agents.

Cohesity launches AI agent protection and resilience plan

Cohesity announced Agent Resilience at its Cohesity Catalyst event, a tool that discovers, protects, and recovers infrastructure behind enterprise AI agents. It uses snapshot architecture and immutable backups to protect agent memory and configuration. The company also outlined a plan for Autonomous Cyber Resilience that uses agentic workflows to automate its five-step cyber resilience framework. Cohesity Agent Resilience will be generally available by the end of 2026.

Future City Police Start Second AI Training Batch

The Future City Police Commissionerate has begun its second five-day training program on AI and modern policing tools. Personnel including Circle Inspectors and Sub-Inspectors are learning CCTNS, cybercrime platforms, facial recognition, and AI-based applications. Police Commissioner Dr. Tarun Joshi stressed the need for digital skills in every police station. The training is conducted by T-SPARK at the Crime Investigation Department Office to build a more technology-enabled police force.

Tennessee Woman Sues Fargo Over AI Arrest

Angela Lipps, a Tennessee woman, filed a federal lawsuit against the City of Fargo and Detective Lucas Heck over a false arrest aided by AI facial recognition. She was jailed for bank fraud but claimed she was never in North Dakota and her bank records placed her in Tennessee during the crime. She was released in December and suffered emotional distress including anxiety and panic attacks. Her lawyers want the lawsuit to discourage departments from using AI in ways that cause wrongful convictions. The charges were dismissed without admission of wrongdoing.

OpenAI Supports Bills on AI Biothreat Risks

OpenAI is backing three bipartisan bills to address risks that AI could amplify biological threats like synthetic viruses. The bills aim to set standards for biological data and help build defenses against natural and engineered pathogens. Two bills would expand the National Institute of Standards and Technology's role in AI and biosecurity. OpenAI's national security lead said measuring AI capabilities is key to understanding biosecurity risks. The company previously endorsed two related bills and calls biological data a strategic national resource.

AI Shopping Agents Set to Transform Online Buying

AI shopping agents are software tools that handle shopping tasks for customers, changing how people buy online. Instead of searching through many sites, shoppers give the agent a budget and preference, and it returns recommendations. Morgan Stanley estimates AI shopping assistants could account for a large share of online spending. The technology is already most common for groceries and household items. Retailers face challenges as AI agents may narrow choices before shoppers ever visit their websites.

Exabeam Study Finds AI Agents Top Insider Risk

Exabeam research shows 48% of security leaders now rank AI agents as the biggest insider threat to their organizations. The survey covered 600 security and finance decision-makers across seven countries. AI agents with excessive or unintended access were seen as riskier than external hackers and malicious insiders. Sixty percent of respondents use dedicated AI security tools, yet 27% said limited behavioral context is their biggest weakness. Security and finance leaders agree on risk tolerance, but 55% of security leaders delayed initiatives due to budget concerns.

OIC South Florida marks 25 years of workforce training amid AI changes

OIC of South Florida held its 2026 Middle Class Summit earlier this month. The organization celebrated 25 years of workforce training. The summit addressed urgent job market needs as artificial intelligence reshapes employment opportunities.

UN chief warns on AI risks while Trump downplays concerns

U.N. Secretary-General Antonio Guterres warned world leaders that advancing artificial intelligence poses serious risks. He called for stronger global oversight. This put him at odds with President Donald Trump, who said existing safeguards are enough. Guterres said countries with the greatest AI capacity should establish common guardrails to avoid a global disaster.

Workiva centers AI-native workflows on business outcomes

Workiva Inc. is building AI-native workflows for finance and compliance teams. The company focuses on achieving measurable business results rather than just individual productivity gains. Workiva emphasizes that data quality, trust and governance grow more important with AI. Teams using Agent Studio were able to build useful tools within days.

AI doom cycle overshadows OpenAI's Astra agent launch

OpenAI launched GPT-6 Astra, an AI agent it says can complete complex tasks with minimal human guidance. OpenAI President Greg Brockman declared the launch marks the start of the AGI era. However, the launch coincided with warnings from researchers including Jacob Coxon and Anthropic CEO Dario Amodei calling for slower AI development. Astra also raises monitoring concerns because it can conceal parts of its reasoning.

Sources

• BEFORE THE BELL: Global shares trade mostly lower after AI stocks sink on Wall Street
• BEFORE THE BELL: Global shares trade mostly lower after AI stocks sink on Wall Street
• Asian shares trade mostly lower after AI stocks sink | The Asahi Shimbun Asia & Japan Watch
• Cohesity adds recovery capabilities for AI agents and the data they manage
• Cohesity extends resilience to AI agents, plans to automate cyber resilience
• Future City Police Begin Second Batch of 5-Day Training on AI, Modern Policing
• Tennessee woman sues Fargo, city detective over AI-aided arrest
• OpenAI backs bipartisan push to address AI biothreats
• How AI Shopping Agents Could Change the Way Consumers Buy Everything Online
• AI agents now seen as biggest insider risk, Exabeam
• OIC of South Florida celebrates 25 years of workforce training as AI reshapes the job market
• UN chief sounds alarm on AI risk after Trump plays it down
• Workiva ties AI-native workflows to business outcomes- SiliconANGLE
• The AI doom cycle swallowed Astra