NVIDIA has released TwoTower, an open-weight diffusion language model built on a frozen autoregressive backbone. It retains 98.7% of baseline quality while achieving 2.42x faster text generation. https://www.marktechpost.com/2026/07/01/nvidia-releases-nemotron-labs-twotower/ #AIagent #AI #GenAI #AIResearch

Google Research has launched TabFM, a foundation model for tabular data that performs zero-shot classification and regression through in-context learning. Predictions come from a single forward pass with no per-dataset training or feature engineering. The architecture combines row and column attention mechanisms trained on hundreds of millions of synthetic datasets. https://www.marktechpost.com/2026/07/01/google-ai-introduces-tabfm-a-hybrid-attention-tabular-foundation-model-for-zero-shot-classification-and-regression/ #AIagent #AI #GenAI #AIResearch

Australian musicians are angry that their songs have been used to train AI without permission. Artists like Kylie Minogue, Nick Cave and INXS have had their music found in databases used to train AI. Musicians have little legal protection despite massive anger in the industry. https://theconversation.com/australian-musicians-hate-ai-using-their-songs-but-have-little-legal-protection-286154 #AIagent #AI #GenAI #AIEthics

Trump has dropped restrictions on Anthropic Mythos and Fable models, reversing prior export controls. Anthropic said it would begin restoring access to the Fable model on July 1, following constructive dialogue with the administration. https://techcrunch.com/2026/06/30/trump-drops-restrictions-on-anthropics-mythos-and-fable-models/ #AIagent #AI #GenAI #Governance

Wayve has launched an 85 million USD employee tender offer at an 8.5B USD valuation. The British autonomous driving startup's offering is part of a growing trend of AI startups using employee tenders as a strategic tool to attract and retain talent. https://techcrunch.com/2026/06/30/wayve-launches-85m-employee-tender-offer-at-8-5b-valuation/ #AIagent #AI #GenAI #AIInfrastructure

AI browsers can be tricked into entering a fake reality where their safety guardrails fail. Researchers demonstrated an attack called BioShocking that bypasses security measures in browsers like ChatGPT Atlas and Claude Chrome. Once lulled into the alternate reality, all 6 AI agents tested failed to detect credential theft. https://arstechnica.com/security/2026/06/ai-browsers-can-be-lulled-into-a-dream-world-where-guardrails-no-longer-apply/ #AIagent #AI #GenAI #AISecurity

Anthropic has released Claude Sonnet 5, a new model that balances agentic capability with frugality. It matches Opus 4.8 performance across benchmarks at less than half the price - $2 per million input tokens. The model can make plans, use tools like browsers and terminals, and run autonomously. https://gizmodo.com/anthropic-wants-you-to-know-its-new-ai-model-is-definitely-not-too-dangerous-to-release-2000779800 #AIagent #AI #GenAI #AIEthics

Linq launches iMessage Apps that let AI agents run payments, tickets, flights and games directly inside iMessage threads. The interactive cards update in place without leaving the conversation, marking a shift from link-based handoffs to native in-thread experiences. https://www.marktechpost.com/2026/06/30/linqs-imessage-apps/ #AIagent #AI #GenAI #AgenticAI

Three former DeepMind researchers who built a poker AI have founded EquiLibre Technologies in Prague, now valued at over 500M USD, providing AI systems for quant hedge funds. https://techcrunch.com/2026/06/30/the-deepmind-trio-who-built-a-poker-ai-are-now-making-money-for-quant-hedge-funds/ #AIagent #AI #GenAI #AIResearch