Meta and Nvidia have renewed the push for open-weight AI models with the release of Meta's Muse Glimmer and Nvidia's Nemotron 3.5 Lightning, both available for free download by developers. This marks a strategic bid to compete against dominant proprietary models from companies like OpenAI and Anthropic, as well as strong offerings from Chinese AI labs. Industry leaders argue that open-weight models can strengthen competition, expand opportunity and mitigate risks by sharing the benefits of AI more broadly. Although there is optimism about the potential for these models in US markets, especially where regulations and security concerns make Chinese models less viable, both Meta and Nvidia need to prove these open-weight models will attract a robust developer base.

AI adoption is accelerating. Readiness is not. New research from Pearson and AWS draws perspectives from learners, educators, and employers across 6 countries to introduce an AI Readiness framework. Read the report now to learn more.

AI is widely adopted but has yet to significantly improve productivity because of management issues, writes Tal Carmi, CIO at WalkMe. Carmi suggests a three-level framework: adoption, budget control and justification. "The CIO's job is to build the layer that makes the right choice the obvious one, and to provide sanctioned alternatives so employees aren't left building their own," Carmi writes.

Meta CEO Mark Zuckerberg has outlined his vision for AI in a 6,500-word essay, advocating for open-source development to prevent concentration of power. Zuckerberg's manifesto envisions AI agents enhancing lives and creating business opportunities. Critics, such as Anthony Aguirre of the Future of Life Institute, warn of the risks of rapid AI development and the potential for human disempowerment.

Signal introduced Automatic Key Verification in 2026, allowing users to confirm their encrypted messages are delivered to the intended recipient without needing to manually compare Safety Numbers. This addresses the longstanding requirement to trust Signal's servers for key distribution by automating verification through a tamper-evident public log and independent auditing. The system runs background checks and displays a green checkmark when verification succeeds, providing users with a seamless and transparent way to ensure message integrity without extra effort.

The SharePoint vulnerability CVE-2026-45659 is a high-severity remote code execution issue that stems from the deserialization of untrusted data. Attackers can exploit this flaw with low privileges and minimal knowledge of the target system, making repeatable attacks possible against unpatched servers. Microsoft released security updates for affected SharePoint versions in May 2026, but hundreds of servers remain exposed. While Microsoft has not yet officially marked the flaw as exploited, CISA has confirmed active abuse by ransomware gangs through its recent updates, emphasizing the importance of immediate remediation.

The vulnerabilities in Zoom were identified by researchers from A Security using AI models that are publicly available. Remarkably, it took less than 20 AI-generated prompts to both discover the bugs and develop a working exploit, demonstrating how AI tools are rapidly lowering the barrier for finding and weaponizing software vulnerabilities.

Enterprises are grappling with severe IT infrastructure shortages, with lead times extending up to 18 months and costs rising by as much as 200%. Hyperscalers are consuming vast amounts of memory, driving up prices for storage systems, servers and networking devices. Experts recommend strategies such as maximizing existing resources, considering alternative vendors and maintaining close communication with finance teams and vendors.

Artificial intelligence (AI) is becoming a routine part of work, but many employees still need training to use it effectively, according to CompTIA's “AI Skills Tracker.” While 80% of professionals use AI regularly and 71% report positive results, organizations are still working to integrate AI into business processes and develop workforce skills. The report highlights growing demand for practical, role-based training focused on security, data analysis and applying AI to real-world business challenges.