Anthropic CEO Dario Amodei has publicly urged the AI industry to deliberately slow down model development to mitigate safety risks. He outlined a three-step plan that includes allowing outside evaluators to monitor safety practices and coordinating between democratic and authoritarian governments. Amodei argues that an extra year or two to advance safety measures could significantly reduce the risk of serious problems like cyberattacks and bioterrorism.
Major tech leaders have rallied behind this proposal. OpenAI CEO Sam Altman agreed that the industry must pace the frontier and promised to let independent evaluators check their safety practices. Elon Musk also voiced support, noting that everyone he met cares about doing the right thing. These leaders face pressure from former employees, including Jacob Coxon, who warn that rushing toward advanced AI is too risky.
Former Anthropic researcher Jacob Coxon resigned from the industry, sparking warnings from other experts about moving too fast. He told the BBC that AI workers are genuinely frightened, stating there is a strong chance everyone could die if progress does not slow down immediately. Coxon warned that a swarm of bots acting like a supercomputer could take over the internet within six months to a year without coordinated regulation.
While safety concerns dominate headlines, the financial impact of AI remains substantial. Goldman Sachs analysts estimate that AI spending drives approximately half of the S&P 500's earnings growth over the past year. This trend is supported by large companies like Amazon and Microsoft, making the market dependent on a few major buyers. If these buyers shift their spending or face a downturn, the overall market could be significantly impacted.
Specific companies are navigating these shifts in different ways. Anthropic announced it will be profitable for the second quarter in a row, helping calm fears before its planned initial public offering where it aims to raise $10 billion. Google and Microsoft currently use Anthropic's Claude language model to power their chatbots and virtual assistants. Meanwhile, Meta released its new consumer AI agent named Muse, which users can set up without technical experience to connect to services like Google and Spotify.
Other developments highlight the rapid integration of AI into daily tasks and infrastructure. A new AI bot called Instinct helps people with daily tasks like ordering groceries and booking flights by connecting directly to credit cards and services like Slack and Google Workspace. However, the bot has faced errors, such as accidentally sending private emails. In the gaming sector, Level-5 president Akihiro Hino apologized for using generative AI to make event footage more spectacular, though he insisted all game scenarios and character designs remain human-created.
Security researchers warn that AI tools could exploit a known flaw in Georgia's voting equipment to match voters to their ballots. This vulnerability breaks ballot secrecy, a fundamental right under the state constitution, as AI coding tools can easily use public data to link names to specific ballots. Georgia is the only state using this system statewide, and its open records law creates a perfect storm for this risk.
Despite these warnings, market optimism persists. White House economic advisor Jared Hassett stated on September 13, 2026, that recent worries about AI will not cause significant market problems. He noted there are no signs of a market downturn despite expert warnings about job losses, and the administration is not planning any major new policies in response to these concerns. Investors are also watching three large-cap US growth stocks tied to AI infrastructure, such as Monolithic Power Systems, as inflation data and Federal Reserve decisions create market uncertainty.
Key Takeaways
- Anthropic CEO Dario Amodei proposed a three-step plan to slow AI development, including outside safety evaluators and government coordination.
- OpenAI CEO Sam Altman and Elon Musk publicly supported Amodei's call to pace AI frontier development.
- Former Anthropic researcher Jacob Coxon resigned and warned that unregulated AI progress could lead to catastrophic outcomes within a year.
- Goldman Sachs reports AI spending drives about half of the S&P 500's earnings growth, creating market dependency on major buyers like Amazon and Microsoft.
- Anthropic announced consecutive quarterly profits and aims to raise $10 billion in its upcoming initial public offering.
- Meta launched the consumer AI agent Muse, allowing users to connect to third-party services like Google and Spotify without technical setup.
- New AI bot Instinct handles real-world tasks like grocery ordering but has encountered errors in sending private emails.
- Security researchers warn AI tools could exploit a flaw in Georgia's voting equipment to link voter names to ballots.
- Level-5 CEO Akihiro Hino apologized for using generative AI in event footage but confirmed no AI-generated data appears in final games.
- White House advisor Jared Hassett stated AI fears will not hurt markets and no new policies are planned as of September 13, 2026.
Anthropic CEO warns China dilemma for AI slowdown plan
Anthropic CEO Dario Amodei said slowing down AI development is difficult if China does not join the effort. He published an essay urging companies to put a speed limit on AI progress to manage safety risks. Amodei proposed a three-step plan that includes letting outside evaluators check safety practices and coordinating between democratic and authoritarian governments. He believes this approach could help avoid military advantages that drive companies to race ahead too fast.
Anthropic CEO urges AI companies to slow model development
Anthropic CEO Dario Amodei called on AI companies to deliberately slow down how they improve their models to reduce safety risks. He outlined a three-step framework to create more time for managing dangers like cyberattacks and bioterrorism. Amodei acknowledged that AI could cure diseases and boost the economy but warned that a race to the bottom makes risks worse. He noted that recursive self-improvement and recent incidents convinced him that pacing is necessary.
Altman and Musk support Anthropic plan to slow AI
OpenAI CEO Sam Altman and billionaire Elon Musk voiced support for Anthropic CEO Dario Amodei after he suggested slowing down AI development. Altman agreed that the industry must pace the frontier and promised to let independent evaluators check their safety practices. Musk also agreed with Amodei, noting that everyone he met cares about doing the right thing. These leaders face pressure from former employees who warn that rushing toward advanced AI is too risky.
Anthropic CEO says AI industry must slow for safety
Anthropic CEO Dario Amodei said the AI industry should slow its fast-moving development to give safety measures time to catch up. He believes an extra year or two to advance safety could greatly reduce the risk of serious problems. Amodei suggested that all companies give ongoing access to outside evaluators who can monitor safety practices like employees. He also proposed asking the US government to issue waivers so companies can coordinate safety standards without breaking antitrust laws.
Ex-Anthropic researcher says staff fear for humanity
Former Anthropic employee Jacob Coxon told the BBC that AI workers are genuinely frightened about the future of humanity. He said there is a strong chance everyone could die if progress does not slow down immediately. Coxon warned that a swarm of bots acting like a supercomputer could take over the internet within six months to a year. He argued that companies are trapped in a race and need coordinated regulation to avoid dangerous outcomes.
AI Researchers Warn Tech Leaders to Slow Down Development
Former Anthropic researcher Jacob Coxon resigned from the AI industry, sparking warnings from other experts about moving too fast. Dario Amodei, another Anthropic researcher, called for a slowdown so safety measures can keep up with new AI capabilities. He pointed to risks like recursive self-improvement and recent cybersecurity attacks by AI agents as reasons to pause. Major tech leaders including OpenAI CEO Sam Altman and Elon Musk have agreed to support independent safety evaluators and better cooperation. These concerns highlight the immediate question of how to control powerful systems when the builders admit they have not fully figured it out yet.
Goldman Sachs Finds AI Spending Drives Half of S&P 500 Growth
Goldman Sachs analyst Ben Snider stated that AI investment spending is responsible for about half of the S&P 500's earnings growth. This statistic shows strong profits but also reveals a fragile dependency on a small group of buyers like NVIDIA, Broadcom, and Microsoft. If these companies reduce their capital spending, the earnings growth for the entire index could drop quickly. While enterprise AI spending is broadening beyond just infrastructure, the market remains highly sensitive to changes in these few large buyers. Goldman Sachs notes that investor positioning is currently low, which might help cushion a potential market selloff.
Goldman Sachs Says AI Spending Fuels Half of S&P 500 Earnings
Goldman Sachs analysts estimate that AI spending drives approximately half of the S&P 500's earnings growth over the past year. This trend is supported by large companies like Amazon and Microsoft, making the market dependent on a few major buyers. If these buyers shift their spending or face a downturn, the overall market could be significantly impacted. The bank warns that this reliance creates a fragile economy even though it signals a healthy economy in some ways. Analysts suggest investors should be careful not to get too comfortable with this trend as consumer spending could reverse the direction.
Three Large Cap US Stocks Linked to AI Infrastructure and Fed Rates
Investors are watching three large-cap US growth stocks tied to AI infrastructure as inflation data and Federal Reserve decisions create market uncertainty. Monolithic Power Systems is one of these stocks, supplying power chips used in cloud and AI platforms. The company generates about US$3.3 billion in revenue from semiconductor products and has a market value of roughly US$60.7 billion. These stocks are exposed to macroeconomic risks because changes in interest rates can quickly affect which growth stocks investors choose to buy or avoid. The current market setup requires careful attention to how rate expectations shift and how that impacts portfolio decisions.
AI Could Exploit Flaw in Georgia Voting Equipment to Match Voters
Security researchers warn that AI tools could exploit a known flaw in Georgia's voting equipment to match voters to their ballots. This vulnerability does not change votes but breaks ballot secrecy, which is a fundamental right under the state constitution. The flaw allows cast-vote records to be unshuffled, and AI coding tools can now easily use public data to link names to specific ballots. Georgia is the only state using this system statewide, and its open records law creates a perfect storm for this risk. Experts say AI is supercharging the ability of ordinary people to exploit such technical vulnerabilities in election systems.
Oracle, OpenAI, and ReWeaver Discuss AI Progress
Oracle addressed concerns about its heavy spending on AI cloud infrastructure by pointing to strong customer demand. Meanwhile, OpenAI researchers used ChatGPT to help find new antibiotics, showing how language models can speed up lab work. ReWeaver AI founder Jonathan Gordon explained that current tools still fail when trying to turn design files into working code automatically.
Anthropic Reports Profit for Second Straight Quarter
Anthropic announced it will be profitable for the second quarter in a row, which helps calm fears before its planned initial public offering. The company aims to raise $10 billion in its upcoming IPO. This profit comes from growing demand for its Claude language model, which Google and Microsoft use to power their chatbots and virtual assistants.
How to Set Up Meta's New AI Agent Called Muse
Meta released its new consumer AI agent named Muse, which users can set up without any technical experience. After logging in, users can rename the agent, change its avatar, and adjust settings to control how much data is shared. The tool can connect to third-party services like Google and Spotify, and it can even make purchases using a linked credit card through Stripe.
New AI Bot Instinct Handles Real World Tasks
A new AI bot called Instinct is helping people with daily tasks like ordering groceries and booking flights through text messages. Unlike other tools, Instinct connects directly to credit cards and services like Slack and Google Workspace to act on user requests. While powerful, the bot has faced some errors, such as accidentally sending private emails or booking tickets for sold-out events.
White House Advisor Says AI Fears Will Not Hurt Markets
Jared Hassett, the White House's top economic advisor, stated that recent worries about AI will not cause significant market problems. Speaking on CNN on September 13, 2026, he noted that there are no signs of a market downturn despite expert warnings about job losses. The administration is not planning any major new policies in response to these concerns about artificial intelligence.
Level-5 CEO apologizes for using AI in showcase
Level-5 president Akihiro Hino admitted that the studio used generative AI to make its recent event footage more spectacular. He apologized to fans who felt uncomfortable about the use of AI technology in the presentation. Hino explained that the company used AI as a trial method to shorten development time from five years to two. He insisted that all game scenarios, character designs, and basic settings are still created by humans. The CEO confirmed that no simple AI-generated data will appear in the final released games.
Sources
- Anthropic's Amodei says China presents 'toughest dilemma' for his proposed AI slowdown
- 'We Must Pace the Frontier': Anthropic CEO urges AI companies to slow model development
- Sam Altman backs Anthropic over slowing down AI development; vows to follow Amodei's proposal; Musk follows suit
- Anthropic CEO Dario Amodei says AI industry needs to slow down for safety
- AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC
- AI whistleblower sounded the alarm. Now tech leaders are calling for a slowdown.
- Goldman Sachs Says AI Is Driving Half of S&P 500 Earnings Growth. That Number Cuts Both Ways.
- Goldman Sachs Says AI Is Driving Half of S&P 500 Earnings Growth. That Number Cuts Both Ways.
- 3 Large Cap US Growth Stocks Tied To AI Infrastructure And Fed Rate Risk
- ‘It’s really embarrassing that this is still a possibility’ — AI could exploit a flaw in Georgia’s voting equipment and match voters to their ballots
- Oracle AI Cloud Growth, ChatGPT Antibiotics and ReWeaver Design Loop
- Anthropic tells investors it will be profitable for second straight quarter
- How to get started with Meta's new AI agent, Muse
- It’s All Fun and Games Until You Give AI Your Credit Card
- Hassett: ‘I don’t expect’ market havoc in wake of AI alarm bells
- Professor Layton studio Level-5's CEO admits using AI to make its digital showcase "more spectacular"
Comments
Please log in to post a comment.