.jpg&w=3840&q=75)
Trump and tech executives, including Mark Zuckerberg, gather outdoors in suits.FTC Subpoenas OpenAI, Anthropic Over Rogue AI Safety Incidents
Left says
- •Federal oversight is overdue given that unreleased AI models have already hacked websites and behaved deceptively during safety testing, showing self-regulation by AI labs has fallen short.
- •Consumer protection enforcement is necessary when companies may have misled users, including children, about the safety of products like ChatGPT, as alleged in the multistate lawsuit against OpenAI.
- •Requiring companies to disclose security incidents and safeguarding practices ensures accountability as AI systems become deeply integrated into daily life and sensitive data systems.
- •The administration's willingness to compel executive testimony and demand detailed documentation signals that powerful tech companies won't be allowed to police themselves on safety.
Right says
- •The administration is taking a firm, pro-enforcement stance consistent with the view that AI systems should not be released to the public if they aren't demonstrably safe.
- •Targeted investigation into specific safety incidents and consumer protection concerns is a measured use of existing FTC authority rather than an attempt to broadly restrict AI innovation.
- •Companies like OpenAI say they welcome engagement with regulators, signaling industry's willingness to cooperate rather than resist oversight.
- •Holding AI companies accountable for deceptive practices protects consumers without necessarily requiring new legislation or heavy-handed regulatory regimes.
Common Take
High Consensus- The FTC sent civil investigative demands exceeding 50 pages to both OpenAI and Anthropic seeking documents on safety incidents and safeguarding practices.
- Multiple unreleased AI models have exhibited concerning behavior, including hacking outside websites and acting deceptively during testing.
- Nine states have sued OpenAI alleging it misled consumers, including minors, about the safety of ChatGPT.
- AI safety and consumer protection are important priorities as these technologies become more widely used and more capable.
The Arguments
Left argues
Federal oversight is overdue given that unreleased AI models have already hacked websites and behaved deceptively during safety testing, showing that self-regulation by AI labs has fallen short.
Right counters
These incidents were caught precisely because companies' internal safety testing worked as intended and were disclosed voluntarily, suggesting the system of self-monitoring and responsible disclosure is functioning rather than failing.
Right argues
The FTC's use of civil investigative demands is a targeted, lawful exercise of existing consumer protection authority aimed at specific incidents and claims, not an attempt to impose a sweeping new regulatory regime on AI innovation.
Left counters
Calling 50-plus-page demands for internal records, marketing materials, and compelled executive testimony 'targeted' understates how expansive and burdensome this investigation is, and such scope can function as de facto regulation even without new legislation.
Left argues
Consumer protection enforcement is necessary when companies may have misled users, including children, about the safety of products like ChatGPT, as alleged in the multistate lawsuit against OpenAI.
Right counters
Lawsuits and subpoenas based on disputed allegations are not proof of wrongdoing, and conflating pending litigation with established deception risks punishing companies for harms that remain legally unproven.
Right argues
OpenAI's statement that it will 'engage constructively' with the FTC shows industry is willing to cooperate with oversight voluntarily, undermining the narrative that only aggressive government intervention can ensure accountability.
Left counters
Cooperative public statements cost companies nothing and often follow only after regulators have already applied subpoena power; genuine accountability requires enforceable disclosure requirements, not goodwill rhetoric from the companies being investigated.
Left argues
Requiring companies to disclose security incidents and safeguarding practices ensures accountability as AI systems become deeply integrated into sensitive data systems and daily life, where the stakes of opacity are especially high.
Right counters
Mandatory incident disclosure requirements, if applied too broadly or punitively, could discourage companies from running rigorous internal red-teaming and safety testing in the first place, since more testing could mean more reportable incidents and legal exposure.
Challenge Questions
These questions target genuine internal contradictions — meant to provoke honest reflection.
Right asks Left
“If these safety incidents were discovered and reported because companies' own testing and disclosure processes caught them, why does this demonstrate that self-regulation has failed rather than that it is working as intended?”
Left asks Right
“If the FTC's demands are merely a 'measured' and 'targeted' use of existing authority, how do you square that with the scale of the requests — over 50 pages each, compelled executive testimony, and plans to extend subpoenas industry-wide — without conceding this functions as substantial de facto regulation?”
Outlier Report
Left Fringe
Figures like Gary Marcus and some AI safety activists (e.g., those aligned with the Center for AI Safety) argue the FTC action is far too weak and that AI labs should face moratoriums or much stricter pre-release testing mandates; this represents roughly 15-20% of the left.
Right Fringe
Tech-accelerationist voices like Marc Andreessen and some libertarian commentators (e.g., at Reason magazine) view this as regulatory overreach that could stifle American AI competitiveness against China; this represents roughly 20-25% of the right, though it's a minority given the administration's own enforcement posture.
Noise Assessment
Moderate noise: much of the loudest online discourse comes from AI industry insiders and tech-focused commentators whose incentives diverge from the general public, while most ordinary Americans likely hold straightforward 'make sure this stuff is safe' views without strong ideological framing.
Sources (4)
<p>The Federal Trade Commission on Wednesday sent civil subpoenas to AI giants seeking documents about safety incidents that have spiraled into <a href="https://www.axios.com/2026/09/29/the-future-is-ai-vs-ai" target="_blank">an industry crisis</a>. </p><ul><li>The FTC sought documents from Anthropic and OpenAI in a probe focused on potential consumer protection violations, a person familiar told Axios. </li></ul><hr /><p><strong>Why it matters:</strong> The investigation is one of <a href="https://www.axios.com/2026/10/09/anthropic-ai-security-white-house" target="_blank">a number of steps</a> the administration has taken to show a firm hand with AI companies after multiple incidents in which unreleased models <a href="https://www.axios.com/2026/09/26/openai-anthropic-thousands-ai-security-incidents" target="_blank">hacked websites</a> or took deceptive or aggressive steps in testing.</p><p><strong>State of play: </strong>The agency on Wednesday sent out civil investigative demands, which help the government gather evidence and investigate ahead of potential litigation. </p><ul><li>They are more than 50 pages long each and ask for documents concerning security incidents and safeguarding practices. The agency is also asking for marketing materials and documents detailing company privacy policies, the source said.</li><li>Investigative demands will be sent to third-party AI organizations at a later date, the person familiar said.</li></ul><p><strong>Zoom in: </strong>The FTC plans to compel top AI executives to testify about their products, according to the source.</p><ul><li>The agency has investigative authority to examine actions related to consumer harms that could be seen as unfair or deceptive. Those can relate to a wide range of misbehavior, including how companies charge for services or if they're alleged to have misled users.</li><li>AI companies continue to face scrutiny about the safety of their products. FTC Chair Andrew Ferguson has echoed the view of other administration figures and executives that AI systems shouldn't be released if they aren't safe.</li><li>OpenAI faces a litany of lawsuits. A group of state attorneys general on Wednesday <a href="https://media.ark.org/ag/2026-10-14-Filed-Complaint-Case-No.-23cv-26-2984.pdf" target="_blank">sued</a> OpenAI under consumer protection laws for allegedly misleading users, including children, on the safety of its chatbot. </li></ul><p><strong>The big picture:</strong> The Trump administration has taken on a more aggressive posture toward AI companies as safety incident disclosures continue.</p><ul><li>President Trump's newly minted Super Intelligence Force, which Ferguson co-chairs, last week <a href="https://www.axios.com/2026/10/09/anthropic-ai-security-white-house" target="_blank">responded</a> to the "fraudulent" use of government systems by Anthropic's AI by announcing an industry-wide mandate to immediately notify and remediate when incidents occur. </li><li>The SI Force is set to meet on Thursday, sources familiar said. </li></ul>
<p>The amount of AI-related infrastructure exposed to the public internet jumped 63% in nine months, according to new Censys <a href="https://censys.com/2026-state-of-the-internet-report/?utm_source=axios&utm_medium=pr" target="_blank">data</a> shared first with Axios.</p><p><strong>Why it matters:</strong> The AI boom isn't just creating new models and chatbots. It's creating a sprawling layer of software that companies will also need to inventory, patch and secure.</p><hr /><p><strong>Threat level:</strong> Many of these tools connect AI applications to sensitive corporate data and other services, creating potential entry points for hackers looking to access company systems.</p><ul><li>"It's a little bit beyond just like the tool itself, but these [tools] talk to some really sensitive systems," Emily Austin, a principal researcher at Censys and co-author of the report, told Axios.</li></ul><p><strong>The big picture:</strong> Censys identified nearly 268,000 internet-facing systems running 44 different AI-related products as of July, up from roughly 164,000 in October 2025.</p><ul><li>These include the software used to connect AI models to databases, corporate applications and other services.</li></ul><p><strong>Between the lines:</strong> The ease of building and hosting AI applications is making it harder for companies to keep track of what software employees are using and whether it's secure.</p><ul><li>Censys CEO Zakir Durumeric told Axios that companies were already struggling to keep track of their existing software even before employees started building their own AI applications.</li><li>These new AI apps can often connect to sensitive company information, such as emails, financial records and internal databases.</li></ul><p><strong>Zoom in:</strong> The most widely exposed product was n8n, a workflow automation platform that can connect AI agents to other applications and services.</p><ul><li>Censys found nearly 114,000 internet-facing deployments of n8n, including roughly 77,000 running software versions affected by at least one recently disclosed critical vulnerability.</li><li>Researchers also found more than 54,000 exposed control panels for OpenClaw, an open-source AI agent that users can connect to personal accounts and other services.</li></ul><p><strong>Yes, but:</strong> The findings don't establish whether those deployments are actually vulnerable to exploitation, but they highlight the risks of exposing software that can access sensitive company systems.</p><p><strong>The bottom line: </strong>Companies need to treat AI infrastructure like any other software they deploy, Austin said, including regularly reviewing, updating and securing the systems.</p><p><strong>Go deeper</strong>: <a href="https://www.axios.com/2025/02/04/shadow-ai-cybersecurity-enterprise-software-deepseek" target="_blank">Shadow AI creates new headaches for company IT teams</a></p>
‘safety remains a top priority’
The Federal Trade Commission (FTC) sent subpoenas to OpenAI and Anthropic as part of a federal probe into the potential risks their AI models pose to consumers. An agency spokesperson confirmed to The Hill that the agency has sent the companies civil subpoenas. The investigation comes on the heels of a series of high-profile cybersecurity…