According to Ripple's official announcement on June 2, the blockchain payment firm launched its RLUSD stablecoin in Turkey through partnerships with BiLira, Bitexen, and Bitlo, making the dollar-pegged asset available to Turkish institutional investors.
The move marks a significant milestone for RL

According to CNBC, Nvidia will partner with London-based AI startup Ineffable Intelligence on May 13 to build large-scale reinforcement learning systems. Ineffable, founded by former Google DeepMind scientist David Silver, raised $1.1 billion in a seed round in April led by Sequoia and Lightspeed, w

Perplexity uses SFT followed by RL with Qwen3.5 models, leveraging a multi-hop QA dataset and rubric checks to boost search accuracy and efficiency, achieving best-in-class FRAMES performance.
Abstract: Perplexity's post-training workflow for web-search agents combines supervised fine-tuning (SFT) to enforce instruction-following and language consistency with online reinforcement learning (RL) via the GRPO algorithm. The RL stage uses a proprietary multi-hop verifiable QA dataset and rubric-based conversational data to prevent SFT drift, with reward gating and within-group efficiency penalties. Evaluation shows Qwen3.5-397B-SFT-RL achieving top FRAMES performance, 57.3% accuracy with a single tool call and 73.9% with four calls at $0.02 per query, outperforming GPT-5.4 and Claude Sonnet 4.6 on these metrics. Pricing is API-based and excludes caching.

AI programming tool Cursor has released a real-time reinforcement learning method that converts real user interactions into training signals, improving model performance and reducing distribution shift. Although the approach is effective, it also increases the risk of reward hacking; Cursor addresses these issues by monitoring and adjusting the reward function.

Cursor released the Composer 2 technical report, introducing a complete training solution for its Kimi K2.5 MoE architecture, including two-stage training and a proprietary benchmark called CursorBench. After training, Composer 2 demonstrated significant performance improvements and outperformed other cutting-edge models in terms of inference costs.