Governance and Regulation
A renewed commitment to strengthening the United Nations for its next era
(Microsoft) Eighty years ago, the first United Nations General Assembly convened in London, marking the start of a new era of global cooperation. Today, the context in which the UN operates has changed significantly. The UN system is being asked to deliver results with greater speed and precision, often amid tightening resources and growing demands. In response, the UN is advancing reforms to become more agile, accountable, and efficient. The UN80 initiative, launched by UN Secretary-General Antonio Guterres, is about making the UN system fit for the future, strengthening its resilience, responsiveness, and capacity for innovation across humanitarian response, development, and international security. – https://blogs.microsoft.com/on-the-issues/2026/01/20/a-renewed-commitment-to-strengthening-the-united-nations-for-its-next-era/
Interpreting Claude’s Constitution
(Kevin Frazier – Lawfare) AI labs, which develop, test, and deploy models for commercial use, have extensive tools at their disposal to shape the capabilities and behaviors of their AI services. Which data a lab relies on to train a model, for instance, may alter its outputs. Likewise a model trained pursuant to one architecture over another may have different characteristics. Anthropic’s creation of a “constitution” for its large language model, Claude, may introduce a new vector for how AI labs, users, and the public shape the capabilities and characteristics of leading AI tools. Released publicly by Anthropic on Jan. 21, “Claude’s Constitution” is in many ways meant to emulate the structure and effect of a traditional constitution. First, it sets forth high-level values to generally inform how the model responds in any given situation without precisely dictating the proper course of action. Second, those values supersede any other instructions, rules, or policies adopted by Anthropic or a user. Third, it is meant to evolve, albeit only when mandated by the circumstances. Just how significant is “Claude’s Constitution”? Answering that question depends on a few factors. There’s the technical factor of the extent to which the constitution alters how a model behaves. There’s the substantive factor of the constitution’s existing provisions and, perhaps, future amendments. And, there’s the external engagement factor of which actors may have a role in shaping the constitution and its implementation. Select comparisons to the U.S. Constitution inform these inquiries and tee up a larger question: what, if anything, does Claude’s Constitution mean for the average user, average American, and average individual the world over? For now, the answer is unclear. If, however, Claude’s Constitution proves effective at shaping model behavior and is subject to a broader discourse among AI governance stakeholders, then January 21, 2026 could well become a key date in AI policy lore. – https://www.lawfaremedia.org/article/interpreting-claude-s-constitution
The AI Preemption Executive Order’s BEAD Strategy Faces Steep Legal Hurdles
(Brian McGrail – Lawfare) On Dec. 11, 2025, President Trump signed an executive order titled “Ensuring a National Policy Framework for Artificial Intelligence” that, as part of a broader attempt to block state regulation of AI, directs the National Telecommunications and Information Administration (NTIA) to condition an estimated $21 billion in broadband funds on states rolling back AI regulations the administration deems “onerous.” But the administration’s chosen legal mechanism faces a significant statutory obstacle: The Broadband Equity, Access, and Deployment (BEAD) Act does not clearly allow NTIA to condition BEAD funding on state AI policy. The state attorneys general preparing to challenge the executive order have the better reading of the BEAD statute—and even if a court found the text ambiguous, interpretive canons would likely break the tie against the administration. The major questions doctrine requires clear congressional authorization before an agency asserts power over questions of vast economic and political significance. Federalism-protecting canons require the same before Congress conditions federal funds on state policy or displaces traditional state authority. Section 5(a) implicates each of these concerns: It would leverage a broadband infrastructure program to reshape national AI policy, targeting state laws governing disclosure, liability, and consumer protection. Yet BEAD—a statute about deploying service and connecting locations—never mentions AI. The statute does not offer the clarity these canons require. – https://www.lawfaremedia.org/article/the-ai-preemption-executive-order-s-bead-strategy-faces-steep-legal-hurdles
Geostrategies
Anthropic CEO Criticises U.S. Approval of AI Chip Exports to China at Davos
(AI Insider) Following the U.S. administration’s decision to approve sales of high-performance AI chips from Nvidia and AMD to approved Chinese customers, debate over AI export controls has intensified. Speaking at the World Economic Forum, Dario Amodei, chief executive of Anthropic, warned that allowing the export of advanced processors used to train AI models posed significant long-term national security risks for the United States. – https://theaiinsider.tech/2026/01/22/anthropic-ceo-criticises-u-s-approval-of-ai-chip-exports-to-china-at-davos/
Security and Surveillance
UK Executives Warn They May Not Survive a Major Cyber-Attack, Vodafone Survey Finds
(Phil Muncaster – Infosecurity Magazine) Major cybersecurity breaches at UK retailers and carmakers last year have raised boardroom awareness of online threats, but many senior executives warn they may go out of business if hit by similar incidents. Vodafone Business polled 1000 senior leaders across businesses of all sizes to better understand their attitudes to cyber risk. Some 89% claimed that big-name breaches at M&S, Jaguar Land Rover (JLR) and other firms last year made them more alert to the potential impact of cyber threats. Yet a worrying 10% admitted their organization would likely not survive a similar incident. – https://www.infosecurity-magazine.com/news/uk-execs-warn-may-not-suruvie/
Over 160,000 Companies Notify Regulators of GDPR Breaches
(Phil Muncaster – Infosecurity Magazine) The number of organizations notifying their GDPR regulator of a data breach surged by 22% to a daily average of 443 in 2025, according to DLA Piper. The global law firm has been analyzing GDPR regulatory activity every year since the data protection regulation came into being in 2018. The past 12 months bucked a long-term trend that has seen average daily notifications plateauing and it’s the first time since 2018 that the figure has exceeded 400, DLA Piper noted. – https://www.infosecurity-magazine.com/news/160000-companies-regulator-gdpr/
Pentest tools left online are allowing hackers to exploit Fortune 500 firms
(Ann-Marie Corvin – Cybernews) Hackers are exploiting intentionally vulnerable penetration testing and security training apps that have been mistakenly exposed to the public internet, giving them access to cloud environments including CloudFlare, F5, and Palo Alto Networks. New research from offensive security research firm Pentera focuses on ten popular training tools, including Damn Vulnerable Web Application (DVWA), OWASP Juice Shop, Hackazon, and bWAPP. – https://cybernews.com/security/pentest-tools-left-online-hacking-fortune-500/
LastPass Warns of Phishing Campaign Attempting to Steal Master Passwords
(Danny Palmer – Infosecurity Magazine) LastPass has urged users to be vigilant about an email phishing campaign which is posing as the password manager application provider in attempt to steal master passwords to takeover accounts. The LastPass Threat Intelligence, Mitigation, and Escalation (TIME) team issued the warning after they became aware of an active phishing campaign which started on January 19. The phoney emails claim to be from LastPass and warn users that they need to take urgent action by clicking the link in the message within 24 hours to backup their password vaults ahead of planned maintenance. – https://www.infosecurity-magazine.com/news/lastpass-phishing-master-passwords/
Frontiers and markets
New AI learning tool bridges communication gaps between autistic and neurotypical people
(News Medical Life Sciences) People with autism have brains that are wired differently. This can make them especially strong in some areas-such as noticing patterns, remembering details, or thinking logically-while making other things like social cues or changes in routine more challenging. There can also be stark differences in the way autistic and neurotypical people communicate, to the point where it may seem like each is using a different language, creating complications from social situations to the workplace. For example, while non-autistic people often depend on nonverbal cues like body language and tone of voice, inferring emotion and intent, some autistic people rely on them less, and might interpret linguistic devices like sarcasm or irony literally. – https://www.news-medical.net/news/20260120/New-AI-learning-tool-bridges-communication-gaps-between-autistic-and-neurotypical-people.aspx
AI-driven drug discovery project launches with €60m budget
(UCL News) The five-year project has a budget of more than €60 million, funded by the European Union and industry partners through the Innovative Health Initiative (IHI). The LIGAND-AI consortium brings together 18 partners across nine countries to generate large open, high-quality datasets of how molecules (ligands) bind to proteins, and use them to train AI models capable of predicting candidate molecules as suitable binders for thousands of human proteins for use in medications. Led by Pfizer and the Structural Genomics Consortium (SGC), the consortium of experts across academia, the life sciences industry, technology companies, and research organisations will investigate thousands of proteins relevant to existing and unmet disease areas including rare, neurological, and oncological (cancer) conditions, to generate open and accessible, high-quality, AI-ready data at scale as a public resource. – https://www.ucl.ac.uk/news/2026/jan/ai-driven-drug-discovery-project-launches-eu60m-budget
OpenAI Introduces AI-Based Age Prediction in ChatGPT to Strengthen Youth Protections
(AI Insider) OpenAI has introduced a new age prediction capability within ChatGPT as part of its efforts to better protect minors using the platform. The feature uses AI to assess behavioural and account-level signals, including stated age, account history, and typical usage patterns, to identify users who may be under 18 and automatically apply stricter content safeguards. – https://theaiinsider.tech/2026/01/22/openai-introduces-ai-based-age-prediction-in-chatgpt-to-strengthen-youth-protections/
Tesla Signals Revival of Dojo With Ambition for Space-Based AI Compute
(AI Insider) Tesla is preparing to restart work on Dojo3, its previously shelved third-generation AI chip, with a renewed focus on space-based AI compute. Chief executive Elon Musk said the decision follows progress in Tesla’s internal chip roadmap, including its AI5 design, and marks a strategic shift after the company wound down its Dojo program earlier this year. – https://theaiinsider.tech/2026/01/22/tesla-signals-revival-of-dojo-with-ambition-for-space-based-ai-compute/
H2O.ai Partners with Singapore’s Certis to Deploy AI for Security and Operations
(AI Insider) H2O.ai and Certis Group signed a memorandum of understanding to scale the deployment of predictive and agent-based AI across Certis’ operational environments. The collaboration will combine H2O.ai’s sovereign and predictive AI capabilities with Certis’ Mozart AI orchestration platform to support situational awareness, risk anticipation, and operational planning across urban security, transportation, and critical infrastructure. Under the agreement, the companies plan to co-develop and validate AI models in live operations, including robotics-enabled systems, and to support workforce readiness through AI governance and ethics-focused training initiatives. – https://theaiinsider.tech/2026/01/22/h2o-ai-partners-with-singapores-certis-to-deploy-ai-for-security-and-operations/
Nexxa.ai Raises $9M Seed Round to Develop AI Agents for Heavy Industry
(AI Insider) Nexxa.ai raised a $9 million seed round led by Construct Capital, bringing total funding to $14 million to scale specialized AI agents for heavy industries including rail, construction, and manufacturing. The company’s Nitro platform deploys multi-agent AI on top of existing industrial engineering software, aiming to modernize complex workflows without replacing legacy systems or disrupting mission-critical operations. Nexxa.ai said the new capital will support expansion of its forward-deployed engineering model and accelerate adoption across core U.S. infrastructure sectors, following early deployments that showed measurable returns in weeks. – https://theaiinsider.tech/2026/01/22/nexxa-ai-raises-9m-seed-round-to-develop-ai-agents-for-heavy-industry/
Microsoft Announces Robotics Model Rho-Alpha Built on Company’s Vision-Language Models
(AI Insider) Microsoft has introduced Rho-alpha, a research robotics model designed to translate natural-language instructions into physical actions for dual-arm robots performing manipulation tasks. Rho-alpha extends vision-language-action approaches by incorporating tactile sensing and is being developed to learn from human feedback during deployment, targeting tasks such as button pushing, knob turning, plug insertion, and tool handling. The model remains in research, with testing on dual-arm industrial and humanoid robots and planned access through an early access program and later via Microsoft Foundry, supported by training that combines physical demonstrations with synthetic simulation data. – https://theaiinsider.tech/2026/01/22/microsoft-announces-robotics-model-rho-alpha-built-on-companys-vision-language-models/
Germany’s NEOintralogistics Raises €3M Seed Round for Robots-As-A-Service Warehouse Automation
(AI Insider) NEOintralogistics closed a €3 million seed round led by the Amadeus APEX Technology Fund with participation from Cetus Holding, backing its robotics-as-a-service (RaaS) platform for warehouse automation. The company offers a pay-per-pick robotic picking system that enables brownfield and greenfield warehouses to deploy automation in weeks with zero upfront CapEx, targeting a market where roughly 80% of warehouses remain manual due to cost barriers. Proceeds will fund commercial scaling, customer acquisition, product refinement, R&D expansion, and team growth, as NEOintralogistics deepens collaborations with partners including Magazino (Jungheinrich), GLS, and BITO. – https://theaiinsider.tech/2026/01/22/germanys-neointralogistics-raises-e3m-seed-round-for-robots-as-a-service-warehouse-automation/
Dam Secure Announces $4M in Funding to Secure AI-Generated Code for Enterprises
(AI Insider) AI security startup Dam Secure raised $4 million in seed funding led by Paladin Capital Group to address security risks from AI-generated code entering production. Founded by Patrick Collins and Simon Harloff, the company is building an AI-native platform that detects logic flaws missed by traditional security tools and enforces security rules across codebases in plain English. The funding will support accelerated product development and go-to-market expansion as enterprises increasingly rely on AI coding tools. – https://theaiinsider.tech/2026/01/22/dam-secure-announces-4m-in-funding-to-secure-ai-generated-code-for-enterprises/
Orbem Secures €55.5M Series B
(AI Insider) Orbem raised €55.5 million in a Series B led by Innovation Industries, with Supernova Invest and existing backers participating, to scale its AI-powered MRI platform. The company has made MRI fast and scalable for industrial use, is already profitable, and leads in-ovo sexing across five European countries with over 170 million eggs scanned. The new funding will support U.S. expansion, new poultry and produce solutions, and longer-term applications of its growing biological dataset in human healthcare. – https://theaiinsider.tech/2026/01/22/orbem-secures-e55-5m-series-b/