Meta readies its “Iris” AI chip for September production

Meta plans to begin manufacturing its custom data-center AI chip, codenamed Iris, in September as part of a four-generation silicon roadmap. Internal documents show the company aims to scale its compute infrastructure to 14 gigawatts by 2027, doubling its 2026 target, with the Broadcom-designed, TSMC-manufactured chip clearing testing in just six weeks. The move is Meta’s most aggressive push yet to reduce its dependence on Nvidia and AMD GPUs. Source: Tech Startups

Helsing raises $1.8 billion in Europe’s largest defense-tech round

Munich-based Helsing, Europe’s leading AI defense startup, closed a $1.8 billion Series E at an $18 billion valuation, with backing from Goldman Sachs Alternatives, Dragoneer, Iconiq, CPPIB, and JPMorgan. The round cements Helsing’s position in autonomous drones and battlefield intelligence, and underscores Europe’s push to build sovereign AI capabilities in sensitive sectors. Source: Tech Startups

Publishers seek court sanctions against OpenAI

The New York Times and other major publishers filed a motion asking a federal court to sanction OpenAI for allegedly withholding key evidence about how its models are trained in their ongoing copyright lawsuit. A ruling on sanctions could shape discovery and settlement dynamics across the many similar cases facing AI developers. Source: Tech Startups

Anthropic hires former Monzo CEO for its compute team

Anthropic has recruited Tom Blomfield, co-founder and former CEO of British fintech Monzo, who is taking leave from Y Combinator to join the company’s AI compute team. The hire continues a run of high-profile additions, following Google DeepMind’s John Jumper and former Tesla AI lead Andrej Karpathy, as frontier labs compete for operational talent to scale infrastructure. Source: Tech Startups

US venture funding hits a record $412.7 billion in H1, with AI taking 86%

PitchBook data shows US venture capital reached $412.7 billion in the first half of 2026 — up 30% on all of 2025 — with AI startups capturing $355.9 billion of it. Seven billion-dollar-plus rounds closed in Q2 alone, while non-AI sectors saw flat or declining activity. Source: Tech Startups

What to watch: whether Meta’s September silicon timeline holds — and how Nvidia responds as its biggest customers become its competitors.

Apple sues OpenAI over alleged trade secret theft

Apple filed suit against OpenAI and two former employees in the Northern District of California, accusing the ChatGPT maker of misappropriating trade secrets to accelerate its push into consumer hardware following the $6.4 billion io acquisition. The filing alleges the scheme operated “at every level, from members of its Technical Staff to its Chief Hardware Officer.” Coverage from CNBC notes the case lands at a delicate moment, as OpenAI prepares for a public listing where active litigation from a $3 trillion company becomes a material disclosure risk.

OpenAI responds: “No interest in other companies’ trade secrets”

OpenAI issued its first response on Saturday, with spokesperson Drew Pusateri saying the company has “no interest in other companies’ trade secrets” and remains “focused on building innovative technology that empowers people everywhere,” per MacRumors. Meanwhile, Fortune reported new details of how talent poaching between the two hardware teams set the stage for the dispute. The split is already reshaping products: the redesigned Siri arriving this autumn will run on Google’s Gemini rather than ChatGPT, under the multi-year deal Apple and Google struck in January.

Meta starts charging for its models with Muse Spark 1.1

For the first time, Meta is charging developers to use one of its own models. Muse Spark 1.1, aimed at coding and agentic tasks, is priced at $1.25 per million input tokens and $4.25 per million output tokens — roughly a quarter of rival flagship rates — with API compatibility for both OpenAI and Anthropic SDK formats, according to The New Stack. Access is US-only at launch, and independent benchmark results are still pending.

What to watch: OpenAI’s formal legal reply, third-party benchmarks for Muse Spark 1.1, and Google’s still-unconfirmed Gemini 3.5 Pro general availability date.

Grok 4.5 gets its first independent report card

Twenty-four hours after launch, Artificial Analysis ranked xAI’s Grok 4.5 fourth on its Intelligence Index, behind Claude Fable 5, GPT-5.5, and Claude Opus 4.8 — while awarding it the best agentic tool-use score of any model on the board. The headline caveat: the model’s hallucination rate jumped from 25% to 54% versus Grok 4.3, meaning it knows more but is more confidently wrong when it errs. First-day reviews suggest it is a strong fit for tool-calling agent workflows, and a risky one where output accuracy is the metric.

Snorkel: Grok 4.5 leads on real professional work

Snorkel AI’s independent evaluation on roughly 2,000 expert-authored GDPval+ tasks put Grok 4.5 at a 29% mean pass rate, ahead of GPT-5.5 (22%) and Claude Opus 4.8 (21%). The lead was concentrated in domains requiring professional judgment — legal work, education, and healthcare documentation. Snorkel’s analysis is the most credible early signal on how the new model handles actual workplace deliverables rather than synthetic puzzles.

SK Hynix rides the AI boom to a record US listing

Memory chipmaker SK Hynix — a key supplier for AI data centers — closed its first Nasdaq session up 13% at $168.01 after raising $26.5 billion, the largest US share sale ever by a foreign company. CNBC reports the proceeds will fund aggressive factory and equipment expansion, with the chairman telling investors that AI-driven memory demand “is enormous.”

Illinois signs a landmark state AI safety law

Governor JB Pritzker signed Illinois’s AI Safety Measures Act into law this week, adding another major state to the growing patchwork of US AI regulation. The Transparency Coalition’s legislative update notes that companion-chatbot rules and training-data transparency remain the most active areas of state lawmaking at 2026’s halfway mark.

What to watch: Google’s Gemini 3.5 Pro is now six weeks past its GA target with no confirmed date — every day of delay raises the stakes for its eventual debut.

OpenAI ships GPT-5.6 and launches ChatGPT Work

OpenAI made GPT-5.6 generally available and unveiled ChatGPT Work, an autonomous agent that produces finished documents, spreadsheets, presentations, and web apps. The model arrives in three tiers — Sol, Terra, and Luna — spanning flagship reasoning to fast, low-cost inference, and ships with Codex built in plus a Plan mode that lets users approve steps before work begins. As 9to5Mac reports, the rollout is a direct answer to Anthropic’s Claude Cowork.

SpaceX moves to acquire Cursor maker Anysphere

Days after its record IPO, SpaceX confirmed its intent to acquire Anysphere, the company behind the popular AI coding tool Cursor, in a deal valued at roughly $60 billion. The move signals how aggressively hardware and space firms are now buying their way into frontier software tooling. Details were tracked by Crunchbase.

Kling AI raises $2.8B for generative video

Kuaishou’s generative-video unit Kling AI closed a $2.8 billion round led by Alibaba, Tencent, and Baidu, lifting its pre-money valuation to $15 billion. The raise underscores how China’s tech giants are consolidating behind homegrown video and multimodal models, per Crescendo AI.

Together AI hits $8.3B valuation

Open-model infrastructure provider Together AI raised $800 million in Series C funding, reaching an $8.3 billion valuation. The platform lets enterprises train and serve open-source models, and the round reflects sustained appetite for alternatives to closed frontier APIs. The deal was among the week’s largest.

Google DeepMind pushes Gemini 3.5 Pro to July 17

Google DeepMind delayed Gemini 3.5 Pro to July 17 for what it describes as an architectural rebuild targeting stronger mathematical reasoning, SVG scene generation, and image quality. The slip keeps GPT-5.6 and Anthropic’s lineup in the spotlight for another week.

What to watch: whether Gemini 3.5 Pro lands on schedule next week, and how ChatGPT Work stacks up against Claude Cowork in real-world agentic tasks.

Assignment context: RSVZ is a dynamic organization where more than 150 people work within the Information Technology department. The organizational culture is informal. RSVZ is a bilingual environment with French-speaking and Dutch-speaking colleagues and stakeholders. We are currently looking for an experienced .NET Integration Developer to expand team Black (consisting of 12 people, including 2 Integration Developers). Team Black is a “platform” team that works transversally within Informatica and provides technical support to the development teams. The team works according to SAFe (Scaled Agile Framework) methodology to deliver high-quality and scalable applications.  The team’s current focus is on making the Azure Cloud platform usable for other development teams. Role: The experienced .NET Integration Developer is responsible for the design, development and maintenance of a robust and scalable integration platform in an Azure environment. This person works closely with other employees of the IT department (developers, testers, product owners and scrum masters) to deliver functionalities iteratively and incrementally.  Key Responsibilities: Design, development and maintenance of the integration platform and reusable .NET components using C# and Azure API Management.  Analyze and solve technical problems. You are responsible for solving bugs and providing technical support to less experienced colleagues within and outside your development team. Contact with external parties, including other social security institutions in the context of connectivity. Design, develop and maintain reusable Azure DevOps CI/CD components using C#, Powershell, YAML pipelines and Bicep files. Collaborate with cross-functional, agile teams within a SAFe environment: you work closely with other developers, testers, architects, product owners and scrum masters to iteratively and incrementally deliver supporting tooling. Participate in PI Planning and other Agile ceremonies. Quality assurance and code reviews: You are responsible for conducting code reviews and ensuring code quality through unit tests, integration tests and other QA methods. Contribute to technical documentation and knowledge sharing within and outside the team. You prepare the technical documentation and check the documentation drawn up by less experienced colleagues. Provide technical leadership and mentor less experienced team members. Continuously improving development processes and software quality. You follow trends and developments in technologies and methodologies in your domain.  Automation and continuous integration by setting up pipelines, unit tests to ensure a stable and functional development, test and production environment. Required experience and competencies: Bachelor or Master degree in Computer Science or equivalent experience.  Minimum 5 years of experience in a similar role. 5 years of experience working with Azure 5 years of experience working according to Agile/Scrum. Behavioral skills: Good communication skills, both oral and written. Being able to clearly explain technical concepts to technical stakeholders. Excellent problem solving and analytical skills. Continuous learning and improvement: a mind set focused on continuous improvement and learning new technologies and methodologies. Collaborative. Adaptability and flexibility: the ability to quickly switch between different tasks and priorities. Dealing flexibly with changes and unexpected obstacles. Excellent technical coaching skills. Ability to work proactively and autonomously Sense of responsibility Customer-oriented Languages: French or Dutch 2nd national language: understanding and basic written and oral communication English: understanding and basic written and oral communication Hybrid working: 5 days working week, of which at least 2 days in the office. Teleworking is possible on the other days.

Required Skills

Enterprise Integration Patterns (Nice to have), Ervaring met Azure en Bicep, Ervaring met OAuth Client Credential Flow, Ervaring met RESTful API design met behulp van de OpenAPI specificatie, GIT, gRPC (gRPC Remote Procedure Calls) (Nice to have), NServiceBus (nice to have), Sterke kennis Azure DevOps CI/CD pipelines, Sterke kennis van .NET, C# en ASP.NET Core, Sterke kennis van Azure API management, Sterke kennis van Azure Service Bus, Windows communication foundation (WCF) (Nice to have)

Practical Information

• Company: SMALS
• Location: Brussels – Hybrid
• Start Date: 27 July 2026
• End Date: 31 December 2026
• Duration: 6 months
• Contract Type: Freelance / Mission
• Application Deadline: 15 July 2026

Version EN Function As a Senior .NET Developer, you play a key role in the design, development and maintenance of robust software solutions based on Microsoft technologies. You bring your technical expertise to guarantee the quality of the code, the sustainability of the solutions and the performance of the systems. You are also a technical mentor to other team members. Responsibilities: Actively participate in the development of software components in .NET (Framework and Core); Design scalable technical solutions in collaboration with the architecture team; Technically supervise team members and promote a culture of excellence; Carry out code reviews and ensure compliance with good development practices; Analyze functional needs with analysts and translate them into technical solutions; Ensure the quality, performance, security and maintainability of applications;Participate in optimizing system performance;Integrate and evaluate new technologies in line with business objectives. Behavioral skills Ability to critically analyze and integrate information; Structured and solution-oriented thinking; Ability to generate viable alternatives and draw relevant conclusions; Technical leadership and collaborative spirit; Resilience and adaptability in the face of technical or contextual changes Language skills: Proficiency in one of the two national languages: FR or NL (level C2*) ·         Very good level of English (level C1*)*Levels indicated correspond to the Common European Framework of Reference for Languages (CEFR), which defines the standards of language proficiency in Europe. We offer you an interesting job in the provision of ICT and Data services that have an impact on society, with many advantages: ·         A good balance between professional and private life; Flexible schedule in the provision of 40 hours per week (7:30 a.m. to 8 p.m. with fixed slots between 9 a.m. and 12 p.m. and 1 p.m. and 4 p.m.); Possibility of teleworking; Workspaces organized according to activities; Easy accessibility via public transport common; Car and bicycle parking in the building; Individual locker; Company restaurant; NL VERSIE FunctieAls Senior .NET Developer speel je een sleutelrol bij het wetwerpen, ontwikkelen en onderhouden van robuste software oplossingen op la base de Microsoft technologiesën. I provide technical expertise, guarantee code knowledge and support the technical mentor of your team.              Verantoordelijkheden Actief deelnemen aan de ontwikkeling van software components in .NET (Framework en Core); analysis of the best practices; Gedragsvaardigheden Kritische informatieanalyse en integratie;Gestructureerd en oplossingsgericht denken;Creatief en het bedenken van alternatieven en het trekken van juiste conclusions;Technisch leiderschap en samenwerking;Veerkracht en aanpassingsvermogen aan technische de contextuale veranderingen. Taalvaardigheden: Vloeiend in een de twee landstalen: FR of NL (level C2*) Zeer goed level Engels (level C1*)*De aangegevenlevels komen overeen met het Europees Referentiekader voor Talen (ERK), dat denormen voor taalvaardigheid en Europa vaststelt. We offer you an interesting job in ICT and Data services that has an impact on society, with many advantages: ·         A good balance between work and private life; Flexible working hours of 40 hours per week (7:30 am to 8:00 pm with fixed periods between 9:00 am and 12:00 pm, 1:00 pm and 4:00 pm); The possibility of teleworking; Workplaces organized according to activities; Good accessibility to public transport transportation; Parking for cars and bicycles in the building; Individual lockers; Company restaurant; ·         Various opportunities for self-development: training and internship opportunities (can be followed outside working hours) via our e-academy self-training platform or via the Microsoft Enterprise Skills Initiative platform.

Required Skills

ASP .NET Web API, Authentication/Authorization (JWT, OIDC, claims), Blazor, C# / .NET (Core, 10), Entity Framework Core, MVC, REST API’s, REST, Swagger/OpenAPI, SQL Server, Testing (unit + integration)

Practical Information

• Company: SMALS
• Location: Brussels – Hybrid
• Start Date: 3 August 2026
• End Date: 31 December 2026
• Duration: 5 months
• Contract Type: Freelance / Mission
• Application Deadline: 16 July 2026

As part of the program to take over a part of the Ministry of Defence’s application landscape, we are looking for consultants to strengthen the internal teams.  The Defence program initially consists of the complete takeover of the operational management of about thirty business applications.  These applications are relatively diverse from a business and technological point of view.  The takeover activities can be broken down as follows:  ·         Handover/Takeover (documentation, application maintenance, development of new features, etc.)  ·         Re-engineering and/or complete rewriting of applications in order to align them with current development standards.    Langues  ·         French or dutch : expert level  ·         English : Advanced level (written)   Responsibilities  ·         Development / Maintenance of business applications  ·         Development of application components/webservices  ·         Contribute in the design of architecture components and building blocks (in close laboration with technical lead or senior software architects)  ·         Participation in technical working groups. Scrum meetings and meetings with the users if needed Assistance in the testing, the technical documentation, the deployment, the evaluation and the reporting (Unit, JIRA)    Experiences  ·         Minimum 7 years of full-time professional Java Development experience is required (Full-stack preferred – backend with passion for frontend)  ·         Familiarity with multi-system integration and complex environments    Competencies:  Technical knowledge and practical experience of required technologies to be able to start development and support directly (below) :  Must have :  Java 17+  Spring 4+ Spring Boot 3+  Frontend (Angular 19+)  REST(ful) Web Services,  Openapi  Test driven development (Junit, Mockito, Cypress)  Maven  Docker  GIT NPM  Experience with relational database and O/R mapping: PostgreSQL, SQL Server…  Able to read and understand BPMN, UML & documentation practices  Familiar with AGILE/SCRUM methodologies, test-driven development and continuous integration  Problem solver, able to quickly understand existing code  Solution and customer oriented  Ability to integrate easily into an existing team.  Nice to have:  Keycloak Hibernate  Lombok  Mapstruct Openshift/Kubernetes  Gitlab CI, Sonar,  Oauth 2.0  Flyway or Liquibase Jira/Confluence Mission & Working hoursWe are based in Brussels (just in front of Midi Station).This will be an initial contract until 30/09/2026 – The mission is long term and can be extended.Possibility of remote work (7h48 per day) with a requirement to be on site for a minimum of 2 days per week.EU nationals only.

Required Skills

Angular, Docker, GIT, Java, JIRA/confluence, Maven, npm, Open API, REST, Software Engineering, Spring, Spring Boot

Practical Information

• Company: SMALS
• Location: Brussels – Hybrid
• Start Date: 27 July 2026
• End Date: 31 December 2026
• Duration: 6 months
• Contract Type: Freelance / Mission
• Application Deadline: 18 July 2026

GPT-5.6 Sol impresses on speed, stumbles on benchmarks

The first detailed reviews of OpenAI’s GPT-5.6 Sol landed, praising markedly faster coding and pricing roughly half that of Anthropic’s Fable 5, while flagging inconsistent results on some public reasoning benchmarks. The model remains gated to a small group of trusted partners under a government-reviewed rollout, so hands-on impressions still come mostly from early API access rather than the general public. Read the review.

Google pushes Gemini 3.5 Pro to July 17 for a full rebuild

Google DeepMind confirmed it is delaying Gemini 3.5 Pro to July 17, abandoning the previous Gemini 2.5 Pro architecture in favor of a ground-up redesign. The overhaul reportedly targets mathematical reasoning, SVG scene generation, and image quality as Google races to match GPT-5.6 and Anthropic’s Fable 5. More on the delay.

Latest intelligence index still favors Claude Opus 4.8

An updated snapshot of the Artificial Analysis Intelligence Index places Claude Opus 4.8 narrowly in front at 55.7%, with GPT-5.5 at 54.8% and Claude Opus 4.7 close behind. The tight 2.2-point gap at the top underlines how compressed the frontier has become, even as new releases arrive almost weekly. See the benchmark.

Meta restructures around AI, cutting thousands of roles

Meta began implementing a major AI-focused restructuring, reducing roughly 8,000 positions while reassigning thousands more staff to AI-centric teams. The move highlights how aggressively large platforms are reorganizing headcount and budgets around generative and agentic AI. Latest coverage.

What to watch: Google’s July 17 Gemini 3.5 Pro launch and any broadening of GPT-5.6 access will likely reset the leaderboard yet again.

Function The employee will be part of a team of 5 people, the “Database Administrators”, which is responsible for the management of the SFPD databases, the support of the development teams for everything relating to the databases and the development of tools for their own needs. You will support the modernization of our infrastructure by piloting the migration of our critical databases to Oracle 26ai. Concretely: Administer and optimize our databases in production and out of production Work closely with the teams applications, infrastructure and security and support them during critical incidents Manage the evolution of configurations according to business needs Define and implement a secure access strategy (IAM principles, roles, permissions, audit) Contribute to the demonstration of NIS2 compliance in the data perimeter Contribute to the design of robust and tested backup and restoration strategies Regularly validate incident recovery plans and guarantee recovery in the event of a disaster Document procedures and keep them up to date Communicate clearly on the risks and solutions It goes without saying that their role is crucial for the proper functioning of all the institution’s business applications, in particular the monthly payment of approximately 2,800,000 pensions. Technical and generic skills In-depth knowledge of Oracle RDMS, DataGuard, RMANIn-depth knowledgeof SQL, PL/SQL, RDSKnowledge of LinuxKnowledge of MS-SQL, MySQL, MongoDB, PostgreSQL, Apache and Tomcat are a plus.Knowledge and practical experience of Agile and ScrumAbility to deal with and resolve problems independently, looking for alternatives and implementing solutions by consulting with colleagues.Ability to work in a team, to make your skills available in order to achieve common objectives.Customer orientation, with a constant search for efficiency and productivity.Management of technical project of limited dimensionThe ability to lead a team is a plus Selection criteria In addition to technical skills, the following qualities will also be evaluated: Ability to work in a team Ability to work under pressure Ability to manage projects of limited dimension Ability to design a database architecture Ability to manage a team of technicians (DBAs) The mission involves a physical presence of at least 2 days per week in the SFP premises. Knowledge of French or Dutch speaking languages.  Active knowledge of one of the two national languages ​​and at least passive knowledge of the other.

Required Skills

Agile / Scrum, Automation tooling (IaC), Database administrator, Database architecture, Linux, SQL, PL/SQL, RDS

Practical Information

• Company: Confidential
• Location: Brussels – Hybrid
• Start Date: 17 August 2026
• End Date: 31 August 2027
• Duration: 13 months
• Contract Type: Freelance / Mission
• Application Deadline: 19 July 2026

Today’s story is split between diplomacy and deployment: the United Nations wrapped its first-ever intergovernmental AI dialogue in Geneva just as OpenAI began widening access to its latest model. Here is what mattered in the last 24 hours.

UN closes its inaugural Global Dialogue on AI Governance

The first session of the UN Global Dialogue on AI Governance concluded in Geneva after two days bringing together all 193 member states alongside industry, academia and civil society. Secretary-General Antonio Guterres warned that the world must not let AI “vibe-code” humanity’s future, pressing for guardrails against autonomous weapons and other catastrophic risks. A second session is slated for New York in May 2027.

ITU’s AI for Good summit opens on its heels

The dialogue hands off directly to the ITU AI for Good Global Summit, running July 7-10 in Geneva alongside the WSIS Forum. Together the events turn the city into a temporary capital for AI policy, with governance, safety benchmarks and equitable access dominating the agenda.

White House voluntary release framework nears

In Washington, the administration is finalizing a voluntary framework with OpenAI, Anthropic and Google that would set technical benchmarks for when a model triggers a pre-release security review and clarify domestic versus foreign access rules. An announcement has been expected this week, and it could shape how the next generation of frontier models reaches the public.

GPT-5.6 general access window begins

OpenAI’s GPT-5.6 is entering its broad-access window, running July 7-14. The rollout arrives as the competitive picture tightens: Anthropic has been gaining on OpenAI in self-reported revenue, and buyers are increasingly favoring efficiency over raw scale.

What to watch: whether the White House framework lands this week and how its benchmarks line up with the multilateral principles taking shape in Geneva.