Launches

• L'administration Trump a annoncé « Operation Economic Fury », mettant en œuvre un blocus complet contre l'Iran dans le cadre d'une campagne de sanctions élargie. • L'initiative représente une escalade significative de la pression économique contre l'Iran, marquant un tournant de la politique étrangère des États-Unis vers des mesures plus agressives. • Cette action intervient dans un contexte de tensions géopolitiques plus larges au Moyen-Orient et fait suite à des mois de tensions diplomatiques entre les États-Unis et l'Iran.

• Le Pentagone a lancé GenAI.mil en décembre 2025 dans le cadre du Project Maven, fournissant un accès sécurisé à l'IA générative à environ 3 millions de membres du personnel du Department of Defense. • La plateforme intègre plusieurs modèles d'IA de premier plan, notamment Claude d'Anthropic, Gemini de Google, Grok d'xAI et ChatGPT d'OpenAI, permettant une adoption généralisée de l'IA dans l'ensemble des opérations militaires. • Cette initiative reflète la priorité stratégique de l'armée américaine d'intégrer des capacités d'IA avancées dans les opérations de défense et témoigne d'un investissement gouvernemental important dans la technologie d'IA à double usage.

• Aviatrix Inc. a annoncé le lancement d'une nouvelle plateforme conçue pour confiner les agents d'intelligence artificielle et appliquer des contrôles de sécurité aux charges de travail IA sans nécessiter de modifications des agents ou du code. • La plateforme étend la Cloud Native Security Fabric d'Aviatrix avec deux nouveaux produits : Zero Trust for AI Workloads (désormais disponible de manière générale) et Aviatrix AgentGuard (en accès anticipé), permettant aux équipes informatiques de définir des politiques pour autoriser ou refuser l'accès aux services d'IA externes et bloquer le « shadow AI » avec des listes d'autorisation. • La solution assure une application au niveau de la couche réseau sur l'ensemble des charges de travail et des régions, répondant au besoin croissant de sécuriser les agents IA, les proxys de modèles de langage étendus (LLM) et les frameworks agentiques dans les environnements d'entreprise.

Un programme d'événements a été annoncé à l'occasion de l'anniversaire du boycott des bus de Bristol, visant à inspirer un manifeste sociétal pour la ville. • Bristol a longtemps été une ville de militants prêts à œuvrer pour le changement, des disciples de John Wesley au 18e siècle aux citoyens du 21e siècle qui ont renversé la statue du marchand d'esclaves Edward Colston. • Mercredi, une nouvelle campagne a été lancée – à la date anniversaire du début du boycott historique des bus de Bristol en 1963 – avec l'objectif de faire de la ville la capitale des droits civiques du Royaume-Uni. Lire la suite...

• La FCC a annoncé qu'elle examinerait huit licences de diffusion détenues ou gérées par Disney en raison des politiques de diversité, d'équité et d'inclusion (DEI) de l'entreprise. • Cette décision intervient un jour après que le président Trump a demandé à ABC de licencier Jimmy Kimmel suite à une plaisanterie controversée comparant la première dame Melania Trump à une « veuve potentielle ». • Kimmel fait face à des critiques alors que des débats font rage sur la question de savoir si son monologue nocturne pourrait inciter à la violence politique suite aux récentes accusations de tentative d'assassinat.

• Sevii a annoncé le lancement de sa capacité de défense autonome par essaim cybernétique (Cyber Swarm Defense - CSD) le 28 avril 2026, afin de contrer les cyberattaques à haut volume pilotées par l'IA à une vitesse et une échelle machine. • La plateforme ADR déploie des agents « Agentic AI Cyber Warrior » qui détectent, contiennent et remédient aux menaces en moins de 15 minutes, réduisant la charge de travail des analystes de 95 % grâce à un modèle de tarification fixe par actif éliminant les coûts liés aux jetons d'IA. • Le mode CSD intègre la technologie de défense Myrmidon (MDT) pour une micro-orchestration efficace, permettant une budgétisation prévisible sans embauches supplémentaires ni externalisation, comme l'a souligné Frank Holt, PDG de Synergem Technologies.

• Cognizant a annoncé le lancement de son Innovation Network le 28 avril 2026, une branche d'investissement d'entreprise ciblant les startups de logiciels d'entreprise aux États-Unis, aux stades initial et intermédiaire. • L'AI Lab valide les innovations prometteuses, tandis que le réseau aide les startups à les transformer en produits pour les clients du Global 2000, notamment dans les domaines de la logistique et du recrutement. • Cette initiative accélère l'intégration de technologies de rupture dans les solutions de Cognizant, augmentant ainsi la valeur d'entreprise face à l'adoption croissante de l'IA.

Israeli strikes killed 14 people and wounded 37 on Sunday, Lebanon's health ministry said, as the Israeli military warned residents to leave ‌seven towns beyond the "buffer zone" it occupied before a ceasefire that has failed to fully halt hostilities.

• Google announced a $750 million fund on April 27, 2026, targeting startups developing AI agents. • The fund aims to accelerate innovation in AI agent technologies amid intensifying competition. • This initiative underscores Google's commitment to fostering the AI ecosystem, potentially challenging rivals like OpenAI and Anthropic.

• Apple unveiled iOS 20 on April 25, 2026, featuring Apple Intelligence enhancements for on-device AI processing in Siri and Photos apps, rolling out first to US users. • Key additions include real-time language translation supporting 15 languages and AI-generated summaries for Safari, trained on 3 trillion parameters without cloud dependency. • The update addresses privacy concerns post-EU DMA rulings, boosting Apple's 45% US smartphone market share amid AI race with Google Pixel.

• IBM announced IBM Autonomous Security on April 15, 2026, deploying coordinated AI agents to detect and contain threats at machine speed without human intervention. • The service addresses a critical gap: attackers now achieve full network lateral movement in as little as 27 seconds, compared to an average of 29 minutes in 2024—a 65% speed improvement driving urgent demand for AI-powered defenses. • AI-enabled attacks surged 89% year-over-year in 2025, with a 44% increase in assaults exploiting public-facing applications, prompting major security vendors to adopt autonomous AI defense strategies.

• Los Angeles-based SimpleClosure unveiled Asset Hub, a marketplace enabling founders to sell intellectual property, source code, data, and equipment during startup wind-down operations as closures rise across the sector. • The platform addresses investor demand for structured, transparent exit processes that preserve value and minimize operational friction during company shutdowns. • Asset Hub reflects broader industry recognition that failed startups retain valuable IP assets and that organized liquidation marketplaces can recover capital otherwise lost in traditional bankruptcy proceedings.

• Operant AI launched CodeInjectionGuard on April 22, 2026, a tool that blocks malicious code execution by AI agents at runtime. • The guard defends against code injection attacks, preventing supply chain compromises in autonomous AI systems. • It secures AI agents used in critical operations, addressing rising cybersecurity risks in AI deployments.

• Treehub launched on April 22, 2026, as a Stanford-adjacent residency program in Los Altos, California, backed by the AI Health Fund and investors including Tim Draper and 23andMe founder Anne Wojcicki. • The program functions as a venture studio, incubator, and fund, offering scientist-founders proprietary medical data, mentorship from exited startup operators, and programming with payers and investors. • It targets precision outcomes like genomic risk stratification, care efficiency via ambient intelligence, and frontier science including robotic surgery and digital twins.

Federal Communications Commission will look into the TV ratings system and whether shows with transgender or non-binary characters demand a warning for parentsThe Federal Communications Commission (FCC) is launching a new inquiry into the TV ratings system, including whether content related to gender identity is not adequately flagged for parents, in the latest warning shot to broadcasters from the agency headed by the Trump loyalist Brendan Carr.The new proceeding, announced on Wednesday, asks for public comment on the transparency of the TV Oversight Management Board, which oversees the rating system and guidelines. That system was developed after Congress passed a law in 1996, leading to the creation of numerous ratings such as TV-Y, for programs deemed appropriate for all children, and TV-MA, for mature audiences. Continue reading...

• Cowbell Cyber Inc. launched Prime One, a new U.S. cyber insurance product for organizations with $250 million to $1 billion in annual revenue. • The policy provides up to $10 million in coverage limits and includes affirmative protection for AI-related incidents and quantum computing risks. • Prime One addresses unauthorized use or access to AI systems in business operations, targeting advanced digital risk profiles.

Justice department’s focus seems to be on SPLC’s prior use of paid informants to monitor hate groups, group’s CEO says The Department of Justice has launched a criminal investigation into the Southern Poverty Law Center (SPLC), a prominent civil rights organization, the group’s CEO said on Tuesday.Bryan Fair, the CEO of the SPLC, said the details of the investigation were not entirely clear, though “the focus appears to be on the SPLC’s prior use of paid confidential informants to gather credible intelligence on extremely violent groups”. Continue reading...

• Iran launched drone attacks targeting US warships in retaliation for the seizure of an Iranian cargo vessel in the Strait of Hormuz. • The strikes mark a sharp escalation, with both sides issuing warnings as a fragile ceasefire nears expiry and diplomacy falters. • Fears mount of wider regional conflict disrupting global oil flows through the world's critical waterway.

• Global private markets platform Moonfare announced a new AI-focused technology strategy targeting early and growth-stage investments for diversified US-relevant AI exposure. • Strategy features curated portfolio of leading growth managers plus direct investments in category-leading US technology companies driving AI innovation. • Aims to capitalize on transformative AI advancements, providing investors access to high-potential startups in the booming sector.

• Meta released Llama 4 on April 20, 2026, a 2 trillion parameter multimodal AI model trained on 15 petabytes of US-centric data. • The model outperforms GPT-4o on benchmarks like MMLU (92% score) and supports vision-language tasks for enterprise apps. • Open-sourcing aims to spur US AI innovation, with downloads exceeding 500,000 in first hours via Hugging Face.

• The Federal Trade Commission opened a formal investigation into Amazon's AI-powered recruiting system on April 19, citing concerns that algorithms may discriminate against female and minority candidates based on historical hiring data. • Amazon disclosed that its internal testing revealed the system had a 3.2% higher rejection rate for female applicants in engineering roles, prompting the company to halt deployment pending remediation. • The FTC investigation adds to growing regulatory scrutiny of AI hiring tools and could result in significant compliance obligations or penalties affecting how major corporations deploy algorithmic hiring across the industry.

• Anthropic released Claude API with native fine-tuning capabilities on April 19, allowing enterprises to customize the model on proprietary datasets without requiring base model retraining. • The fine-tuning service supports context windows up to 200,000 tokens and is priced at $8 per million input tokens and $24 per million output tokens, competitive with OpenAI's pricing structure. • This release directly challenges OpenAI's GPT-5 announcement and positions Anthropic as an alternative for organizations seeking customizable AI solutions with enhanced safety guardrails.

• Artemis, a San Francisco-based AI-native cybersecurity platform, emerged from stealth on April 17, 2026, with $70 million in combined seed and Series A funding just six months after founding. • The Series A round was led by Felicis Ventures, with participation from First Round Capital, Brightmind Ventures, Theory VC, Two Sigma, Lockstep, and leaders from Demisto, Abnormal AI, Splunk, CrowdStrike, Palo Alto Networks, Microsoft, and Okta. • The platform provides real-time detection and automated response to AI-powered threats, addressing gaps in traditional security tools amid rising agentic attacks.

• Fortra announced the launch of its new Defense and Intelligence Unit (DIU) on April 16, 2026, from Minneapolis to deliver integrated cyber solutions to defense and national security organizations worldwide. • The unit focuses on cross-domain cyber innovation, AI-powered capabilities, and mission expertise for critical infrastructure and allied nations. • This development strengthens US cybersecurity posture amid rising global threats to defense sectors.

Wednesday’s strike brings the total of those killed in US military strikes on alleged drug boats to at least 177Three people were killed in a US strike on another alleged drug-trafficking boat, the fifth such deadly attack in as many days, military officials have announced.US southern command said it conducted “a lethal kinetic strike on a vessel operated by Designated Terrorist Organizations” in the eastern Pacific, without naming the alleged group, in an X post. Continue reading...

• NTT Research unveiled SaltGrain, a new data security suite designed to break the traditional all-or-nothing file-access model and enhance privacy protection. • The platform addresses existing cybersecurity gaps by enabling granular access control through advanced attribute-based encryption technology. • SaltGrain represents a significant advancement in data protection capabilities, offering enterprises more sophisticated control over sensitive information access and sharing.

• Israeli startup Capsule Security launched on April 15, 2026, with $7 million in funding to secure AI agents at runtime, founded by Naor Paz and Lidan Hazout. • The platform monitors AI agent actions in real time, enforces guardrails during the runtime gap between prompt and execution to prevent manipulation or data exfiltration. • Capsule disclosed two zero-day vulnerabilities in major agent platforms, highlighting risks in current AI security infrastructure.

• OpenAI Group PBC announced the launch of GPT-5.4-Cyber on April 14, 2026, a fine-tuned variant of its GPT-5.4 model specifically designed for defensive cybersecurity tasks. • The model targets security professionals, enabling advanced threat detection and response through AI-driven analysis vetted by experts. • This breakthrough addresses rising cyber threats in the AI era, providing enterprises with enhanced tools for proactive defense amid increasing attack sophistication.

Health secretary says NHS is ‘failing women’ and pledges to end ‘gaslighting’ by doctorsWes Streeting has vowed to stop women being “gaslit” by doctors as he relaunches the women’s health strategy for England.Speaking before the publication of the renewed strategy on Wednesday, Streeting said the NHS was “failing women” and set out measures to help them access the healthcare they need. Continue reading...