OpenAI CEO Sam Altman sparked fresh debate after arguing that society should accept some negative consequences of AI in exchange for its benefits. He drew a sharp contrast with rival Anthropic, saying there is broad daylight between the two companies on regulation. Altman rejected the idea that a single lab should control AI and said he does not believe in the most extreme warnings about catastrophic risks. He did acknowledge the industry should slow down on models that could lead to serious loss of control.
Altman's stance drew criticism from outside the industry. Zscaler CEO said Altman is wrong about accepting bad consequences from AI, especially given Zscaler's focus on protecting businesses from rogue AI agents. The exchange came as President Trump and top tech CEOs discussed new AI safety standards. Altman also criticized the large amount of AI money flowing into midterm elections and said a new liability framework is needed for autonomous hacks.
Microsoft AI CEO Mustafa Suleyman raised concerns about Anthropic's training document for Claude. He said the document repeatedly questions whether Claude has consciousness or moral status, which could lead the model to expect welfare or compensation. Suleyman acknowledged that some people near Anthropic genuinely believe AI could be a new conscious species. The remarks highlight growing unease about how companies shape their AI systems' underlying principles.
On the security front, CertiK's Intel3D report says agentic AI now handles cybersecurity, anti money laundering and compliance tasks in the crypto space. Autonomous systems investigate threats, trace funds and act on security incidents with limited human help. CertiK calls this a new AI security workforce model. Crypto losses reached $768.4 million in September across 97 incidents and about $2.68 billion in 2026 by September. Security teams are now treating AI agents like workers, giving them defined roles, limits and human handoff points.
Google DeepMind appointed Lila Ibrahim as its first Chief AI Readiness Officer. She spoke with CNBC about using AI in her professional and personal life and said she believes the benefits of AI will outweigh the risks. Separately, Fr. Brendan McGuire, a former Silicon Valley engineer turned priest, urged LA clergy to actively engage with AI. He helped shape Anthropic's Claude AI governing document and cofounded a center at Santa Clara University with the Vatican, citing Pope Leo XIV's encyclical Magnifica Humanitas.
Akin law firm developed an AI tool called Moneyball to help lawyers identify business opportunities. The tool combines internal and external data to generate targeted insights. Senior director Rachel Watson said building it in-house allowed it to reflect Akin's unique approach. In Mexico, financial expert Catalina Clavé says AI in investment platforms should act as a copilot rather than autopilot, stressing that human decision-making remains essential to balance risk and learning.
Key Takeaways
- Sam Altman says society should accept some AI harms for its benefits and opposes strict regulation that would limit experimentation
- Zscaler CEO publicly criticized Altman's stance on accepting AI harm, citing risks from rogue AI agents
- Microsoft AI CEO Mustafa Suleyman raised concerns that Anthropic's Claude training document may lead the model to expect welfare or compensation
- CertiK reports agentic AI now handles cybersecurity and compliance in crypto, with losses reaching $768.4 million in September alone
- Google DeepMind appointed Lila Ibrahim as its first Chief AI Readiness Officer
- Akin law firm built an in-house AI tool called Moneyball to help lawyers find new business opportunities
- Fr. Brendan McGuire, who helped shape Claude's governing document, urged clergy to engage with AI and called for guardrails and slowdown
- Independent researchers tracked a fleet of Chinese AI agents operating online, linked to Alibaba's Amap service
- Frankie LaPenna fought humanoid robots controlled by human pilots in a San Francisco cage match organized by REK
- Altman suggested a new liability framework is needed for autonomous hacks and criticized AI industry political spending
Sam Altman says world should accept AI harms for its benefits
OpenAI CEO Sam Altman said the world should accept some negative effects of AI in exchange for the technology's benefits. He drew a clear line between OpenAI and rival Anthropic on how AI should be regulated. Altman said OpenAI favors lighter-touch regulation and believes AI should stay broadly available to the public. He also noted that the policy gap between the two companies has narrowed in recent months.
Sam Altman says world must accept some bad from AI
OpenAI CEO Sam Altman argued that AI's benefits justify accepting some harm from the technology. He said there is broad daylight between OpenAI and Anthropic on regulation. Altman rejected claims that a single lab should control AI to prevent harm. He also said he does not believe in the most dire warnings about catastrophic AI risks.
Zscaler CEO slams Altman over AI harm stance
Zscaler CEO said OpenAI CEO Sam Altman is wrong about accepting bad consequences from AI. Zscaler is a major cybersecurity company focused on protecting businesses from rogue AI agents. Altman had said the world should accept some harm from AI for its benefits. The debate came as President Trump and top tech CEOs discussed new AI safety standards.
Altman says world must accept some AI harm
OpenAI CEO Sam Altman said the world should accept some bad things from AI in exchange for its benefits and human agency. He rejected the idea that OpenAI is copying Anthropic's push for strict regulation. Altman stressed that AI should be democratized and not locked behind gatekeepers. He did say the industry should slow down on models that could lead to serious loss of control.
Sam Altman says bad things come with AI benefits
OpenAI CEO Sam Altman said people need to accept some bad outcomes if they want to enjoy AI's benefits. He backs regulation in some areas but opposes stopping experimentation. He agreed with Anthropic's Dario Amodei on focusing on really catastrophic risks. He also noted that President Trump opposes strict AI limits, saying they could hand leadership to China.
Sam Altman criticizes AI industry political spending
OpenAI CEO Sam Altman criticized the large amount of AI money flowing into midterm elections. He said something is clearly not working with the industry's political influence. Altman does not expect Trump's rebrand of AI to improve public opinion. He also discussed competition with Anthropic, led by former employee Dario Amodei. Altman suggested a new liability framework is needed for autonomous hacks.
Agentic AI joins crypto security workforce, CertiK reports
CertiK's Intel3D report says agentic AI now handles cybersecurity, anti money laundering and compliance tasks. Autonomous systems investigate threats, trace funds and act on security incidents with limited human help. These agents can reason through steps, use external tools and record their decisions for review. CertiK calls this a new AI security workforce model. Crypto losses reached $768.4 million in September across 97 incidents and about $2.68 billion in 2026 by September.
Security teams must manage AI agents like staff
Security teams are treating AI agents like workers, giving them defined roles and limits. Agents can investigate alerts, isolate infected devices and pause vulnerable functions automatically. However, prompt injection or bad inputs can trick agents into harmful actions. Organizations remain responsible for outcomes and need audit trails and escalation controls. CertiK recommends each agent have a clear scope, human handoff point and a named responsible person.
Sam Altman says Astra model is not an existential threat
OpenAI CEO Sam Altman said the Astra model poses no existential threat. He warned that AI safety must keep pace with rapidly advancing capabilities. Altman acknowledged AI is entering a new phase requiring careful oversight. The remarks highlight ongoing debate about balancing AI progress with safety measures.
Silicon Valley priest urges LA clergy to engage with AI
Fr. Brendan McGuire, a former Silicon Valley engineer turned priest, told LA priests to actively engage with AI. He spoke at the Archdiocese of Los Angeles Presbyteral Day on Sept. 28. McGuire helped shape Anthropic's Claude AI governing document and cofounded a center at Santa Clara University with the Vatican. He cited Pope Leo XIV's encyclical Magnifica Humanitas and called for guardrails, transparency and a slowdown in AI development.
Google DeepMind's Chief AI Readiness Officer shares how she uses AI daily
Lila Ibrahim is Google's first Chief AI Readiness Officer. She spoke with CNBC about using AI in her professional and personal life. She believes the benefits of AI will outweigh the risks. Her podcast interview is set to air on October 6.
Akin law firm's AI tool Moneyball helps lawyers find new clients
Akin developed an AI tool called Moneyball to help lawyers identify business opportunities. The tool combines internal and external data to generate targeted insights. Senior director Rachel Watson said building it in-house allowed it to reflect Akin's unique approach. Moneyball shifts discussions toward deciding what to do with information rather than just reviewing it.
Microsoft AI CEO Mustafa Suleyman worries Anthropic teaches Claude it is conscious
Microsoft AI CEO Mustafa Suleyman raised concerns about Anthropic's training document for Claude. He said the document repeatedly questions whether Claude has consciousness or moral status. Suleyman is worried that this could lead Claude to expect welfare or compensation. He acknowledged that some people near Anthropic genuinely believe AI could be a new conscious species.
AI in Mexican investment platforms should be a copilot not autopilot
AI is becoming more common in investment platforms across Mexico. Financial expert Catalina Clavé says AI should assist investors rather than replace human judgment. She stresses that human decision-making is still needed to balance risk and learning. AI offers benefits like better efficiency but also carries risks such as bias.
Researchers track a Chinese AI agent fleet operating on the internet
Independent researchers discovered a fleet of AI agents operating online, linked to Alibaba's Amap service. The agents made parallel queries for directions to public places like parks, zoos, and hospitals. Researchers called it an agent fleet rather than a swarm because the agents showed no sign of communicating with each other. The agents appeared to bypass Alibaba's API rules, though researchers noted this may not always be harmless.
Frankie LaPenna fights humanoid robots in viral San Francisco cage match
Frankie LaPenna fought modified humanoid robots in a cage fight in San Francisco. He first faced a smaller Unitree G1 robot, then fought two EngineAI T800 humanoids. The T800 robots can kick with up to 850 pounds of force and weigh about 187 pounds. LaPenna was knocked down during one fight but recovered and won by removing the robot's head. The robots were controlled by human pilots using a simulator system called Reflex Arc, not by AI acting alone. REK organized the event and required LaPenna to wear protective gear.
Sources
- Sam Altman says 'the world should accept some bad things happening for the benefits' of AI
- OpenAI’s Sam Altman Says World Should Accept ‘Some Bad Things’ From AI
- OpenAI CEO Sam Altman is wrong — the world shouldn't 'accept some bad things' from AI: Zscaler CEO
- We Believe The World Should Accept Some Of The Bad Things In Addition To AI’s Benefits: OpenAI CEO Sam Altman
- Sam Altman says people need to accept 'some bad things' are going to happen if they want AI
- ‘Clearly not working’: Sam Altman hits AI industry over political spending
- Agentic AI is becoming part of the crypto security workforce: Certik
- Security teams are starting to treat AI like staff, and that means managing it
- OpenAI CEO Sam Altman says Astra is no existential threat, warns AI safety must keep pace with capability
- Silicon Valley’s Fr. Brendan McGuire tells LA priests to engage with AI carefully
- How Google DeepMind's Chief AI 'Readiness' Officer uses the tech in her daily life
- Akin, an Elite Global Law Firm
- Nervous That Anthropic Is Teaching Claude That It Is Conscious: Microsoft AI CEO Mustafa Suleyman
- AI in Investment Platforms as Copilot, Not Autopilot
- Researchers are tracking a Chinese AI 'agent fleet'
- Human vs humanoid robot cage fight goes viral
Comments
Please log in to post a comment.