Modelos, agentes, investigación y negocio de la IA.
NeuralTrust, a Barcelona-based startup, has raised €17.2 million in a seed round for its security and governance platform for enterprise AI agents. The firm claims it is the largest seed funding round in cybersecurity in the European Union to date.
Faced with rising AI API fees and limitations on building new data centers, a trend is emerging of migrating AI workloads from the cloud to local environments. Users report significant savings by processing millions of tokens daily with modest hardware, raising questions about the future of the cloud inference market.
Vercel introduced 'The Agent Stack', an integrated set of tools (AI SDK, AI Gateway, Workflow SDK, and Sandbox) designed to build, deploy, and scale AI agents in production. The proposal aims to eliminate the complexity of integrating multiple providers and managing failures by offering a unified infrastructure layer.
Startup Subquadratic claims to have overcome the fundamental limitation of language models: quadratic computational cost. With its SubQ model, it promises 56x faster speeds and reduced energy consumption. However, lack of public access and use of borrowed weights raise doubts.
General Atlantic is negotiating an investment of more than $2 billion in Kling AI, Kuaishou's video division, valuing the company at $18 billion. The deal comes amid rising geopolitical tensions and warnings from Beijing for its AI firms to reject US capital.
A European retail association has requested that AI-generated ads be exempt from upcoming EU transparency rules. The petition, reported by Reuters, seeks to prevent retailers from having to label advertising content created using artificial intelligence, sparking debate on balancing innovation and consumer protection.
Disney is preparing to launch a beta tool in July that will generate TV ads using artificial intelligence, including scripts, video, and music. The decision comes at a time when the public is showing fatigue towards AI-generated content, known as 'slop', and raises questions about quality, authenticity, and employment in the advertising sector.
Andera, a Lightspeed-backed startup, has closed a $37 million Series A round to automate internal audits with artificial intelligence. The platform promises to reduce manual work by 80%, challenging the Big Four's hourly billing model.
OpenAI has published a manifesto in which its CEO, Sam Altman, compares the emergence of artificial intelligence to electricity in the 1920s, and announces the start of a 'third phase' focused on building systems that benefit everyone. The document seeks to distance itself from total automation and emphasizes collaborative development.
AI agents are beginning to operate under legal contracts that define responsibilities, boundaries, and consequences. This approach, analyzed at the O'Reilly conference, promises to transform AI adoption in sectors like finance, healthcare, and critical infrastructure.
Senator Bernie Sanders has introduced a bill that would require AI companies to transfer 50% of their ownership to a public fund, with returns distributed as $1,000 annual dividends to every citizen. The proposal, which has unexpected support from Vice President JD Vance, opens a debate on how to distribute the benefits of artificial intelligence.
Microsoft has revealed a vulnerability called AutoJack affecting its artificial intelligence agents. By chaining three minor flaws, an attacker can achieve remote code execution (RCE) simply by having the agent visit a malicious website. The company has released patches, but the incident underscores the inherent security risks of autonomous agents.