Model

• Startup Genesis AI unveiled an advanced AI model that empowers robotic hands to perform complex physical tasks, backed by a $105 million funding round as reported by TechCrunch on May 7, 2026. • The breakthrough shifts AI from software to physical applications, targeting transformations in manufacturing and logistics sectors. • This development highlights the accelerating trend of embodied AI, with potential to automate intricate manual processes previously limited to human dexterity.

• A*STAR researchers in Singapore created CLUST-McMs, a specialized AI model that beats GPT-4 in summarizing multilingual news from Southeast Asia. • Using a curated dataset of regional news, the smaller model achieved superior accuracy across three evaluation metrics, delivering concise English summaries from overlapping articles in multiple languages. • The study emphasizes localization and targeted data over larger general-purpose models, showing higher fidelity to original sources.

• India's Department of Financial Services Secretary has raised alarms over Anthropic's new AI model, stating it has amplified global cybersecurity concerns. • The development has drawn attention from banking regulators worldwide, prompting calls for immediate action to mitigate potential risks from this AI advancement. • This warning underscores the escalating threats posed by advanced AI in financial sectors, urging regulatory bodies to prioritize safeguards amid rapid technological changes.

• Singapore's Monetary Authority (MAS) met with chief executives of major financial institutions to discuss AI-related cyber threats, prompted by Anthropic's limited release of its Mythos model. • The Cyber Security Agency of Singapore issued a letter urging critical information infrastructure owners to commission board-level reviews of their cyber-risk posture in response to emerging AI security concerns. • The coordinated regulatory response underscores Singapore's proactive stance on managing AI-driven cybersecurity risks in the financial sector and critical infrastructure.

Agreements with Microsoft, Google DeepMind and xAI focus largely on recognizing cybersecurity, biosecurity and chemical weapons risksThe US government has struck deals with Google DeepMind, Microsoft and xAI to review early versions of their new AI models before they are released to the public.The Center for AI Standards and Innovation (CAISI), part of the US Department of Commerce, announced the agreements on Tuesday, saying the review process would be key to understanding the capabilities of new and powerful AI models as well as to protecting US national security. These collaborations will help the federal government “scale (its) work in the public interest at a critical moment”, the agency said in a press release. Continue reading...

• Google, Microsoft, and xAI joined OpenAI and Anthropic in providing the US Commerce Department pre-release access to evaluate their AI models, creating the first voluntary oversight mechanism for all five major frontier AI labs. • The Commerce Department announced the expansion on Tuesday, establishing a framework with no statutory authority and fewer than 200 staff members to address national security concerns before public model releases. • The Mythos crisis forced the US government to confront the absence of formal mechanisms to evaluate powerful AI models that could threaten national security before public deployment.

• Oracle released a new customer security advisory on April 30, 2026, preempting threats to AI models. • The advisory highlights a cybersecurity harbinger, urging updated security postures for AI deployments. • It signals proactive measures against novel vulnerabilities targeting generative AI systems.