The Future of AI, Automation & Digital Innovation
At its Ship 2026 conference, Vercel unveiled a complete set of tools for building, deploying, and operating AI agents in production. From the open-source framework Eve to the secure connectivity system Connect, the company aims to standardize agent infrastructure, directly competing with cloud giants.
The Rokarolla malware has been detected attacking over 200 banking and cryptocurrency apps on Android. It uses overlay techniques and accessibility service access to steal credentials. Users are advised to update their system and avoid installing apps outside Google Play.
France has announced a third phase of the Tibi program, channeling €13 billion from institutional investors into tech startups. The model, which does not use public money directly, aims to strengthen the country's technological sovereignty and serve as a benchmark for Europe.
Despite advances in child safety, Apple is criticized for not implementing system-level measures and instead delegating responsibility to third-party app developers. Richard Pursey, CEO of SafeToNet, points out that this approach does not reach the source of harm and poses a significant risk to children.
The United States has made quantum computing a strategic state priority. Donald Trump signed two executive orders establishing the development of a quantum computer capable of scientific research by 2028 and advancing the migration to post-quantum cryptography in the federal government to 2031, four years earlier than the previous deadline.
The UN urges artificial intelligence companies to disclose the environmental cost of their systems, including carbon, water, and land use. The move aims to curb secrecy in a sector whose energy growth threatens global climate goals.
Intel and AMD have jointly announced the ACE (Advanced Compute Extensions) for the x86 instruction set, designed to accelerate matrix operations (matmul) with high efficiency. This move unifies AI acceleration in desktop and server CPUs, reducing reliance on GPUs for certain workloads.
A federal judge in California ruled that Workday must face a class action lawsuit accusing its AI-driven hiring system of discriminating against job applicants. It is the first case of its kind to reach trial and could set a precedent for the entire HR software industry.
IBM has joined OpenAI's Daybreak Cyber Partner Program, launching a service that uses frontier AI models to identify and confirm software vulnerabilities faster than traditional methods. The alliance aims to integrate advanced artificial intelligence into enterprise security operations, reducing false positives and response times.
Startup Coherence Neuro has implanted its device in three people to detect brain cancer in real time. The system, which combines monitoring and electrical stimulation, opens a new avenue for treating gliomas without invasive surgery.
GitHub has announced significant improvements to Copilot for VS Code: prompt caching and lazy tool search reduce token usage, while the Auto system selects the most suitable model for each task. This brings Copilot closer to a more autonomous and cost-effective AI assistant.
The Five Eyes intelligence alliance (US, UK, Canada, Australia, New Zealand) issued a joint warning: the next generation of artificial intelligence (frontier AI) could unleash unprecedented cyber threats within months. The report urges governments and businesses to take urgent action to bolster defenses.