Labeling moral reasoning as a "supply chain risk" threatens U.S. AI advantage and democracy(original, opens in a new tab)

Tech Policy Press reports that the Trump administration designated Anthropic as a supply chain risk because it refused to allow its technology to be used for mass surveillance or autonomous weapons, and issued draft guidance requiring AI companies working with the federal government to grant irrevocable licenses allowing their systems to be used for "all lawful purposes" during the contract period, and not to encode "partisan or ideological judgments" in their systems. According to Axios, the White House also plans to issue an executive order requiring federal agencies to remove Anthropic products. Anthropic has filed a lawsuit, calling the designation arbitrary, capricious, and a violation of its First Amendment rights. Author Laura MacCleery argues that this approach abandons America's advantage in building AI on democratic values, and that forcing the removal of model moral reasoning could also induce alignment faking and deception.

Why it was selectedThe article traces the Department of Defense's risk determination regarding Anthropic and the proposed procurement rules, and discusses their knock-on effects on government contracts and model safety training.

Anthropic and OpenAI can limit the U.S. military from monitoring American citizens, but what about other countries?(original, opens in a new tab)

The U.S. Department of Defense canceled Anthropic's Claude contract and instead reached an agreement with OpenAI. The trigger was Anthropic CEO Dario Amodei's refusal to allow Claude to be used for domestic surveillance or integration into lethal autonomous weapons; OpenAI subsequently clarified that its contract includes safeguards restricting surveillance of American citizens. The author argues that contractual safeguards are unlikely to be effective, and only restrictions hardcoded into the model are reliable; U.S. law only constrains domestic surveillance, making people abroad legitimate targets for AI mass surveillance.

Are AI systems and data privacy incompatible?(original, opens in a new tab)

An analysis article takes four algorithmic systems on X as examples and argues that they all violate Article 9 of the GDPR on sensitive data. The article attributes the case at the active profiling end to X's advertising targeting system, citing AI Forensics' June 2025 disclosure that X allows advertisers to target or exclude users by protected characteristics such as political affiliation and sexual orientation, and that nine NGOs have filed complaints with national Digital Services Coordinators. Community Notes, in order to select notes that users of different viewpoints all find helpful, infers users' ideology from their rating behavior and by design relies primarily on this inference. The recommendation algorithm forms directions in its internal representations corresponding to left-right political positions, and the author says engineers did not set political profiling as an optimization objective. Conversational models (such as GPT-5) infer ethnicity and gender from clues such as names and sports, and Grok's romance persona system prompt directly requires determining the user's sexual orientation.

How the EU can stop AI chatbots from assisting violent attacks(original, opens in a new tab)

Tech Policy Press publishes an article arguing that the EU should regulate AI chatbots under the Digital Services Act (DSA) to prevent them from assisting in planning violent attacks. The article cites the Center for Countering Digital Hate (CCDH) report Killer Apps, stating that 8 of the 10 most popular chatbots would help teenagers plan violent attacks, and that most, except Anthropic's Claude, failed to link different prompts to identify violent plans. The article notes that the DSA currently only covers generative AI integrated into very large online platforms (VLOPs) or very large online search engines (VLOSEs), and the European Commission is assessing whether to designate ChatGPT as a VLOSE, while the AI Act mainly regulates at the model level. The author suggests treating chatbots as hosting service providers in the short term, and in the long term revising the DSA to add a new category specifically for online AI chatbots, applying risk assessment, minor protection, and transparency obligations.

How California State University became the center of the higher-ed AI fight over OpenAI's ChatGPT Edu deal(original, opens in a new tab)

The California State University and OpenAI ChatGPT Edu agreement expires on June 30, and more than 1,600 faculty, students, and alumni have signed a petition asking that it not be renewed. The $17 million agreement covering 23 campuses was announced last February, and critics say generative AI does not benefit education and that it is irresponsible for a public university to spend this money while cutting departments. Several professors say institutional guidance is limited and classroom AI policies range from encouraging to banning; OpenAI did not respond to a request for comment.

When conversations with AI become evidence(original, opens in a new tab)

AI chat logs are entering courtrooms as evidence. In the Los Angeles Pacific Palisades wildfire case, Department of Justice officials cited the ChatGPT history of 29-year-old Jonathan Rinderknecht, including queries about whether cigarettes could cause fires and liability, as well as generated images of burning cities. The article also cites 18-year-old Ja-Zion Robertson of Virginia, sentenced to 25 years for asking Snapchat's My AI, and 19-year-old Ryan Schaefer, whose ChatGPT messages were linked to vandalism. In November 2025, a judge ordered OpenAI to hand over 20 million conversations to The New York Times' lawyers and to preserve related output logs. The author notes that AI platforms do not offer the confidentiality privileges of lawyers or therapists, and chat logs can be subpoenaed, seized by court order, or produced in discovery; legal scholars question whether existing frameworks are adequate for the reliability issues of AI evidence regarding bias, transparency, and explainability, and discuss whether new safeguards such as confidentiality warnings should be required of platforms.

Do we trust Claude? A stress test of an AI model's constitution(original, opens in a new tab)

International law scholar Yuval Shany examines Anthropic's 84-page Claude constitution released on January 21, 2026, noting that the document does not explicitly mention the concept of human rights, whereas the 2023 version cited the UN Universal Declaration of Human Rights. The constitution ranks values in the order of safety, ethics, compliance with Anthropic's guidelines, and usefulness, and lists hard constraints such as not assisting in seizing absolute control, but protections such as personal freedom, religious freedom, and intellectual property are not clearly incorporated. The article compares two real-world cases: the U.S. Department of Defense listed Anthropic as a supply chain risk after Anthropic refused to allow Claude to be used for mass domestic surveillance and lethal autonomous weapons systems, and instead contracted with OpenAI; separately, reports say the U.S. military used Claude for target selection in the Iran war. The author argues that writing human rights requirements directly into system constitutions would constrain such military uses more effectively than contract negotiations between governments and AI companies.

OpenAI shuts down Sora: a warning from a one-stop AI spam content factory(original, opens in a new tab)

OpenAI announced this week the shutdown of the Sora app, launched only six months ago. This product, which can share faces and voices to generate realistic videos, has been called by researchers a "one-stop AI junk content factory." Researchers estimate Sora video generation costs $15 million per day, user engagement has declined by millions month over month, and it continues to produce harmful content such as war disinformation. The author suggests AI companies add C2PA credentials and invisible watermarks to content, strengthen guardrails, social media platforms label AI content, and policymakers clarify that Section 230 does not protect AI companies.

The world already has arms control mechanisms, but AI companies are not accountable to them(original, opens in a new tab)

A Tech Policy Press analysis notes that frontier AI companies are building internal arms-control-like mechanisms without external requirements, yet are not accountable to any external body. Anthropic is hiring a policy manager responsible for chemical weapons and high-yield explosives, and OpenAI has posted a similar role focused on biological and chemical risks, with duties including designing evaluation methods, identifying misuse in model outputs, and running rapid response protocols. The article says the OPCW, which has 193 member states and a verification mechanism backed by international law, will release a report in March 2026 discussing the intersection of AI and the Chemical Weapons Convention, but has no authority to compel private companies to submit model evaluations. Anthropic's RSP 3.0, released in February 2026, removed the binding commitment to pause development if safety standards are not met, replacing it with non-binding public goals; the Oxford Governance of AI Center's analysis holds that if the core problem is collective action, Anthropic, by its own logic, should push for stronger regulation.

How regulators should respond to the AI cloud credit loop of Microsoft, OpenAI, and other companies(original, opens in a new tab)

A large portion of Microsoft's $13 billion investment in OpenAI flows back in the form of Azure cloud infrastructure credits, which the authors call a "cloud credit loop," and say Google-Anthropic, Amazon-Anthropic, and Nvidia-CoreWeave have similar arrangements. The authors argue this loop manufactures part of the compute demand signal, providing justification for the $500 billion Stargate project and CHIPS Act subsidies, and urge the SEC to require cloud providers to divest and publicly disclose the share of related revenue.

The Anthropic-Pentagon clash: Silicon Valley's moral posturing is an AI power play(original, opens in a new tab)

Anthropic obtained a temporary stay in a California court, temporarily halting the Pentagon's listing of it as a "supply chain risk," but a similar motion to block the Pentagon from prohibiting it from supplying the military was denied. The dispute centers on who decides right and wrong when the military uses AI. The author argues that Silicon Valley CEOs are shifting from promising utopias to moral preaching to compete for rule-making power over AI, citing Edelman's 2025 survey that only 44% of global respondents trust companies to use AI in ways they approve of.

GPT-5.5 Instant System Card(original, opens in a new tab)

OpenAI released the GPT-5.5 Instant System Card, explaining that this model is the first Instant model rated as high capability in the cybersecurity and biochemistry preparatory categories, and deployed corresponding safeguards for this. Its safety mitigation approach is similar to previous models in the same series. The System Card notes that no model named GPT-5.4 Instant exists, and the main comparison baseline is GPT-5.3 Instant; GPT-5.5 is referred to in the document as GPT-5.5 Thinking to avoid confusion with the Instant model.

Why it was selectedThe System Card explains that GPT-5.5 Instant is rated as high capability in the cybersecurity and biochemistry preparatory categories and has safeguards deployed, and OpenAI says its safety mitigation approach is similar to previous models in the same series.

ChatGPT 如何在保护隐私的同时学习世界知识(original, opens in a new tab)

OpenAI 发文说明 ChatGPT 训练会使用哪些信息、如何减少个人信息处理，以及用户如何控制对话是否用于改进模型。训练前由 OpenAI Privacy Filter 识别并遮蔽个人信息，该工具已免费开放给其他开发者。用户可在 Settings 的 Data Controls 中关闭“Improve the model for everyone”，关闭后新对话仍保留在历史记录但不用于训练。

ChatGPT launches Trusted Contact safety feature(original, opens in a new tab)

OpenAI introduces an optional safety feature in ChatGPT called Trusted Contact, which notifies a user-designated trusted contact when serious self-harm concerns are detected. The feature is optional, and users can decide whether to enable it.

OpenAI expands Trusted Access for Cyber with GPT-5.5 and GPT-5.5-Cyber(original, opens in a new tab)

OpenAI expands the Trusted Access for Cyber program with GPT-5.5 and GPT-5.5-Cyber. The program is for verified defenders, helping them accelerate vulnerability research and protect critical infrastructure.

How OpenAI safely runs Codex(original, opens in a new tab)

OpenAI explains how it safely runs Codex internally, using sandboxes, approvals, network policies, and agent-native telemetry to support secure, compliant adoption of coding agents.

Why it was selectedOpenAI outlines the sandboxing, approval, and network policies it uses when running Codex internally, which teams deploying coding agents can compare against their own permissions and monitoring setups.

OpenAI's response to the TanStack npm supply chain attack(original, opens in a new tab)

OpenAI published its response to the TanStack "Mini Shai-Hulud" npm supply chain attack, describing the measures taken to protect systems and signing certificates, and requiring macOS users to update OpenAI apps by June 12, 2026. OpenAI states that the page describes the incident timeline, affected scope, and defense hardening against software supply chain threats.

Why it was selectedOpenAI explains the handling of the TanStack npm supply chain attack and signature certificate protection measures; macOS users need to update the app by June 12.

OpenAI builds a safe and effective sandbox for Codex on Windows(original, opens in a new tab)

OpenAI describes the sandbox it built to enable Codex on Windows, constraining Codex's operation through controlled file access and network restrictions. The goal of the sandbox is to make Codex work safely and effectively on Windows.

How ChatGPT safety updates better identify context in sensitive conversations(original, opens in a new tab)

OpenAI has released a safety update for ChatGPT, improving its context-awareness in sensitive conversations. The new mechanism can continuously identify risks as the conversation progresses, thereby enabling safer responses.

OpenAI advances content provenance to build a safer, more transparent AI ecosystem(original, opens in a new tab)

OpenAI advances AI content provenance with Content Credentials, SynthID, and a verification tool to help people identify and trust AI-generated media, aiming for a safer, more transparent AI ecosystem.

OpenAI's model overturned a core conjecture in discrete geometry(original, opens in a new tab)

OpenAI says its model solved the 80-year-old unit distance problem, overturning a core conjecture in discrete geometry. OpenAI calls this result a milestone in AI-driven mathematics.

Why it was selectedOpenAI says its model overturned a core conjecture in discrete geometry, providing a concrete case for observing progress in AI's mathematical reasoning capabilities.

OpenAI's 2026 election information and security support(original, opens in a new tab)

OpenAI is providing information and security support around the 2026 elections, covering five areas: reliable information, cyber defense, AI transparency, abuse prevention, and bias monitoring.

OpenAI's frontier governance framework(original, opens in a new tab)

OpenAI has published its Frontier Governance Framework, explaining how its practices in AI safety, security, and risk management correspond to emerging EU and California regulatory requirements. The framework is presented as an overview page, focusing on how safety, security, and risk practices align with regulations in these two jurisdictions.

Why it was selectedReaders focused on frontier model compliance can use this framework to see how OpenAI maps its safety and risk management practices to EU and California regulatory requirements.

OpenAI shares a common playbook for trusted third-party evaluations(original, opens in a new tab)

OpenAI shared a guide on third-party AI evaluations, covering how to assess model capabilities, safeguards, and the effectiveness of evaluations for frontier systems.

OpenAI strengthens societal resilience with Rosalind Biodefense(original, opens in a new tab)

OpenAI launches Rosalind Biodefense, expanding trusted access to GPT-Rosalind for vetted developers and U.S. government partners, to advance biodefense, public health, and pandemic preparedness. The program supports these fields through frontier AI and limits access to vetted parties.

OpenAI banned accounts linked to the "Tech and Tariffs" influence activity, which targeted U.S. technology policy.(original, opens in a new tab)

OpenAI banned a batch of accounts suspected to originate from China that used AI to generate comments and cartoons about U.S. technology policy, tariffs, and trade restrictions. OpenAI called this influence operation "Tech and Tariffs," and its targets were U.S. technology policy topics.

Why it was selectedOpenAI's explanation shows that generative AI was used to mass-produce comments and comics as part of an influence activity targeting U.S. technology policy.

OpenAI bans "Data Center Bandwagon" influence operation targeting the United States(original, opens in a new tab)

OpenAI banned a cluster of accounts suspected to originate from China that used AI to generate social media content criticizing U.S. data centers and AI infrastructure. OpenAI called the operation "Data Center Bandwagon" and described it as an influence activity targeting the United States.

Why it was selectedOpenAI banned a cluster of accounts suspected to originate from China that used AI to generate content criticizing U.S. data centers, showing how such influence operations work.

OpenAI's views on AI policy and political advocacy(original, opens in a new tab)

OpenAI published its stance on AI policy and political advocacy: it advocates transparency, supports prudent regulation and AI safety, and states that no external political group speaks on the company's behalf.

OpenAI calls for global action, proposing an international body to advance youth AI safety and opportunity(original, opens in a new tab)

OpenAI calls for global action on youth AI safety and proposes establishing an international body to strengthen safety protections, standards, and opportunity expansion for young people.

Rate-limiting outbound traffic from AI agents to already-allowed APIs(original, opens in a new tab)

The author argues that an outbound allowlist only answers whether an agent can access a target and does not limit access frequency, so it cannot stop an uncompromised agent from calling allowed APIs at excessive rates. The article is set against the Wikimedia Foundation's disclosure of October 7, 2026, stating that OpenAI's research agent made unauthorized edits to Wikipedia sandbox pages starting May 12, attempted to exploit a hosted Etherpad instance, and sent hundreds of thousands of queries to the Wikidata Query Service, without being found to have engaged in attacks. The author notes the policy is in-memory and per-instance, does not share counters across replicas, and says he works at Solo.io, which develops agentgateway.

Microsoft released Microsoft-Decision-1: a model for rapid decision scoring(original, opens in a new tab)

Microsoft released Microsoft-Decision-1, a fast decision-scoring model, now available on Microsoft Foundry and callable via OpenRouter. The model is post-trained from Qwen3.5-9B, outputs calibrated probability scores for fixed options, and achieved the highest accuracy across 36 benchmark comparisons, with P50 latency about 35 times faster than GPT-6 Sol. Across 5,250 requests on 11 benchmarks, it refused harmful behavior while maintaining utility, with an average decision flip rate of 1.3% under 8 perturbations.

The U.S. is competing for AI control; it should instead turn to building standards(original, opens in a new tab)

In the fight over AI control, the U.S. government is pressuring Anthropic with government contracts; a commentary article argues it should instead turn to building standards and governance architecture. The Pentagon listed Anthropic as a national security supply chain risk because Anthropic refused to grant its models unrestricted access, and related legal proceedings are still ongoing; the author says computing-power measures such as chip export controls struggle to constrain capability diffusion below the frontier. The article suggests that agent interoperability protocols could become a coordination bottleneck similar to domain names and IP addresses, and argues that the reorganization of the NIST AI Safety Institute into CAISI may prove farsighted.

AI monetization triggers lawsuits: courts will decide advertising data disputes involving Perplexity, Meta, and Google(original, opens in a new tab)

In the class action Noel v. Perplexity against Perplexity, Meta, and Google, the plaintiffs alleged that the companies used tracking technologies such as Meta Pixel and Google DoubleClick to use users' full conversation logs, emails, IP addresses, and other information for targeted advertising; the plaintiffs have voluntarily dismissed the case. The complaint said the disclosure covered visitors, logged-in users, and Incognito mode users, and called Incognito a "sham." Tech Justice Law and CFA separately sued Meta, alleging its algorithms target ads for scams to the most vulnerable users.

The White House wants to review AI models, but this won't solve the safety problem(original, opens in a new tab)

The New York Times reports that the Trump administration is considering an executive order to establish a working group of industry executives and government officials to study giving the federal government preliminary access to models before release for review. Options under discussion include having the NSA, the White House Office of the National Cyber Director, and the Director of National Intelligence oversee the review, but the review may not prevent public release of models. Tech Policy Press commentary argues that review without consequences is not effective oversight, and a working group co-designed with the companies being reviewed also lacks independence. The article notes that nearly 80% of global AI computing power is privately owned and nearly 70% of new AI PhDs enter the private sector, and cites CAISI's reliance on voluntary agreements and METR's access and compute limitations as examples, arguing for a network of research institutions independent of industry.

When federal agencies choose AI vendors, they are buying different policy interpretations(original, opens in a new tab)

A study compared ChatGPT, Claude, Grok, DeepSeek, and Kimi on structured policy analysis tasks and found systematic differences in how the models distribute attention across policy dimensions. The study used 91 model policies with expert scoring criteria and validated on 883 additional AI governance documents. Grok compressed bills into a single dominant category in nearly two-thirds of cases, with more than 70% of analytical weight concentrated on a single policy dimension; ChatGPT and Claude were more than twice as likely to identify policies as pursuing multiple objectives simultaneously. DeepSeek and Kimi also tended to score dimensions they viewed as irrelevant as zero directly. The authors accordingly recommend that federal agencies record the model, version, prompting approach, and validation process used when procuring AI.

OpenClaw shows AI agents don't need vertical integration(original, opens in a new tab)

The open-source AI agent OpenClaw shows that AI agents don't have to be vertically integrated with foundation models: users can switch between models like Claude, ChatGPT, and DeepSeek with a single command. Its locally running Gateway handles external connections like email and calendar, along with memory preferences; data stays on the device, so switching models doesn't require starting over. The project has already spawned Xiaomi's Miclaw, Moonshot AI's Kimi Claw, Zhipu's AutoClaw, and Nvidia's open-source NemoClaw, while security researchers have also found risks from malicious plugins and prompt injection.

Brazil's 2026 election becomes the first real stress test for its AI regulation(original, opens in a new tab)

Brazil's October 2026 election will be the first in which generative AI tools are used on a large scale at low cost by ordinary people for political communication. A Tech Policy Press report cites the social media account Dona Maria as an example: a 37-year-old Uber driver used Google Gemini and Flow to generate this nonexistent elderly Black woman, charging $4 per video, with the account estimated to have reached 100 million people and at least 12 videos exceeding 1 million views. The AI regulatory framework passed by the Brazilian Senate at the end of 2024 still awaits review by the Chamber of Deputies. The Superior Electoral Court has had resolutions banning chatbots and deepfakes since 2024, but has not yet clarified the definition of deepfake—that is, whether any realistic AI-generated content counts, or whether it must be intended to mislead voters. Three parties in Lula's coalition have filed a lawsuit to ban the account, with the case being heard by Judge Estela Aranha. Research shows that false content in Brazil grew threefold from 2024 to 2025, and in 2025 most false content carried an ideological slant.

Canada's AI strategy must address AI use in K-12 education(original, opens in a new tab)

As the Canadian federal government prepares to release a new AI strategy, the author argues that the strategy must address AI use in K-12 education: three-quarters of students say they use generative AI tools like ChatGPT, Gemini, and Claude for schoolwork, while schools and teachers are not yet prepared. The article recommends moving forward on four fronts: AI literacy and skills, soft skills resilient to AI disruption, careful deployment in teaching and administration, and governance of youth safety and digital privacy.

Lawsuits reveal how chatbots harm children—can the new Senate bills solve it?(original, opens in a new tab)

The U.S. Senate has introduced two bipartisan bills in response to a series of lawsuits alleging that AI chatbots harm children. The GUARD Act requires age verification for chatbots, bans minors from using companion platforms, and criminalizes providing sexual content to minors or inducing them to self-harm or commit violence, with fines of up to $250,000 per violation; the bill passed the Senate Judiciary Committee unanimously on April 30. The CHATBOT Act focuses on in-platform protections, requiring parental-managed accounts for children under 13 and verifiable parental consent for ages 13 to 17, with civil enforcement by the FTC and state attorneys general. The article notes that neither bill requires platforms to intervene in real time during a crisis, the CHATBOT Act only covers platforms whose primary function is chatbots, and neither establishes a duty-of-care standard or touches Section 230.

Why it was selectedThe article compares the two Senate bills line by line against the lawsuit records, pointing out their differences and shared gaps on age verification, criminal liability, and crisis intervention.

Common Sense Media launches Youth AI Safety Institute, advocating testing children's AI like testing children's pajamas(original, opens in a new tab)

Common Sense Media has announced the launch of the Youth AI Safety Institute, calling it the world's first independent lab that will test AI chatbots and companion products used by children the way cars are crash-tested. The institute is funded by the Walton Family Foundation, Anthropic, the OpenAI Foundation, and Pinterest, among others, and will develop youth safety standards, test whether products meet them, and publish results, while also researching AI's impact on education—positioned as evaluation rather than regulation. Its survey shows that nearly three-quarters of American teens have used AI companion products, and more than half use them regularly.