May 19, 2025: Microsoft Unveils Windows AI Foundry for Local Development - Microsoft has launched Windows AI Foundry to enhance local AI development on Windows devices. Announced at Microsoft Build 2025, the service integrates large language models optimized for different hardware types, facilitating seamless AI integration. Windows ML serves as the built-in inference runtime, simplifying model deployment by providing new APIs for AI tasks and supporting LoRA for fine-tuning LLMs alongside advanced search capabilities.

The Azure AI Foundry also received updates, focusing on streamlined model selection and multiagent app deployment. It now offers an affordable Developer Tier for experimentation, making it easier for developers to experiment and deploy models efficiently. These updates aim to improve the accessibility and functionality of AI tools on Windows and Azure platforms.

May 19, 2025: Microsoft Unveils AI Platform for Rapid R&D - Microsoft introduced Microsoft Discovery, an enterprise AI platform to speed up scientific R&D, at Build 2025. This agentic AI system improves hypothesis generation and experimentation using a graph-based knowledge engine and specialized AI agents, quickly leading to a data center coolant prototype discovery in under four months.

Companies like GSK and Este Lauder are early adopters, using it for drug development and personalized skincare. Running on Nvidia's Azure AI infrastructure, Discovery aims to enhance scientific innovation with speed, accuracy, and scale.

May 19, 2025: Grok 3 AI Joins Microsoft Azure Platform - Microsoft has teamed up with Elon Musk's xAI to offer Grok 3 and Grok 3 mini via Azure AI Foundry, ensuring managed access with standard service-level agreements. Grok, noted for its edgy and permissive characteristics, has stirred controversy over sensitive content and unauthorized modifications.

The Azure-hosted versions aim to mitigate these issues, offering tighter restrictions and improved data integration, customization, and governance features. These enhancements set them apart from the more controversial models available on Musk's social network, X.

May 16, 2025: OpenAI Debuts Codex AI for Coding Tasks - OpenAI introduces Codex, an advanced AI coding agent within ChatGPT, powered by the codex-1 model. Codex generates cleaner code and manages multiple tasks at once, boosting software engineers' productivity. Initially, it is available to ChatGPT Pro, Enterprise, and Team subscribers, with plans to expand to Plus and Edu users.

Codex operates in a secure cloud and connects to GitHub, enabling users to perform tasks from bug fixes to feature development. OpenAI aims to monetize Codex by setting rate limits and selling additional credits.

May 16, 2025: Liquibase Enhances Databricks Lakehouse Management - Liquibase has launched the Liquibase Pro Extension for Databricks, enhancing schema changes in Databricks SQL for quicker and more secure data operations. This integration aids data teams managing analytics and AI workloads by decreasing reliance on manual SQL scripts and notebooks, which hinder delivery and elevate risk. The extension offers features like version-controlled deployment and changelog validation to ensure speed, compliance, and reduced complexity.

Databricks representative Ariel Amster emphasizes the significance of consistent change management as organizations adopt lakehouse architecture. The extension is now available for Databricks users looking for improved governance and change automation.

May 16, 2025: Codex AI Enhances ChatGPT for Coders - OpenAI has launched Codex, a coding-optimized AI integrated into ChatGPT, powered by the codex-1 model for programming. Codex assists developers by generating and reviewing code, with response times from one to 30 minutes. It operates in isolated containers without web access and can be customized via an AGENTS.md file to match production environments.

Codex achieved a 75% accuracy rate in tests. Initially available as a research preview at no additional cost, it will offer flexible pricing and more integrations, including an updated Codex CLI using the lightweight o4-mini model.

May 16, 2025: OpenAI Introduces Parallel-Tasking Codex AI Agent - OpenAI has launched a research preview of its Codex AI software engineering agent, introducing parallel tasking capabilities. Built on a fine-tuned version of OpenAI's o3 reasoning model, this cloud-based agent autonomously handles tasks such as writing features, fixing bugs, and proposing pull requests.

Designed for ChatGPT Pro, Enterprise, and Team users, Codex operates in a secure, isolated environment, enhancing security and collaboration. As competition grows in the AI coding tools space, OpenAI highlights Codex's role in transforming developer workflows and boosting productivity.

May 16, 2025: AWS Launches Strands SDK for Easy AI Development - Amazon Web Services has released the Strands Agents SDK, an open-source toolkit aimed at streamlining AI agent development. Initially used by AWS teams, it expedites the transition from prototype to production. The SDK supports multiple language models like Amazon Bedrock and Anthropics API and integrates with third-party applications. It offers over 20 prebuilt tools and facilitates multi-agent collaboration.

Strands SDK helps reduce infrastructure costs by managing AI data efficiently. It is already employed in various AWS services, including Amazon Q Developer and AWS Glue.

May 16, 2025: Cohere Acquires Ottogrid for AI Market Research Integration - AI startup Cohere has acquired Ottogrid, a Vancouver-based platform specializing in automated market research tools. Ottogrid will cease its standalone product, with its features and team integrating into Coheres platform. This integration particularly enhances North, a ChatGPT-style application for knowledge workers.

The acquisition aligns with Coheres strategic shift towards private AI deployments, following a significant revenue shortfall in 2024. Ottogrid, initially founded as Cognosys in 2023, raised $2 million and offered AI-powered document analysis before joining Cohere.

May 16, 2025: Theom Secures $20M to Revolutionize Data Governance - Theom, an AI-native data operations platform, raised $20 million in Series A funding led by Wing VC, with Sentinel One, Snowflake, and Databricks also participating. These funds aim to enhance product development and expand the team to address increasing demand for secure, AI-driven data governance across cloud and Gen AI environments.

Theom's platform automates sensitive data discovery, monitoring, and policy enforcement, redefining traditional governance and security. Its integration with major platforms such as Snowflake and Databricks underscores trust in its capability to secure data in an AI-driven future.

May 16, 2025: NTT DATA Unveils Smart AI Ecosystem - NTT DATA has introduced an enterprise-grade Smart AI Agent Ecosystem with industry-specific solutions to transform businesses. This ecosystem includes a patented solution that upgrades legacy bots into autonomous agents, enhancing decision-making in sectors like healthcare, finance, and manufacturing. Strategic partnerships, including one with OpenAI, bolster innovative AI services.

Their approach promises improved performance and cybersecurity while unlocking new operational capabilities. The initiative aims to guide clients through AI adoption with responsible, scalable solutions tailored to specific industry needs.

May 15, 2025: Kore.ai and AWS Enhance Enterprise AI Integration - Kore.ai has enhanced its partnership with AWS by integrating its AI platforms with AWS services such as Amazon Bedrock. This strategic agreement aims to boost AI adoption in enterprises, leveraging AWS's robust infrastructure to improve customer experience and operational efficiency.

As an AWS Innovation Award winner, Kore.ai joins the AWS ISV Accelerate Program, collaborating with AWS sales teams to broaden its market reach. This partnership enables businesses worldwide to access advanced AI technologies easily through the AWS Marketplace, driving transformative outcomes.

May 15, 2025: Google Enhances Android and Chrome with New AI Features - Google enhanced Android and Chrome with AI-driven features, notably updating TalkBack to interact with Gemini on image and screen content, and introducing Expressive Captions to capture speech nuances. These enhancements improve accessibility for visually impaired users by describing visuals and adding sound labels.

Chrome now uses OCR for scanned PDFs to enable screen reader interaction and text manipulation. Chrome's Page Zoom on Android allows customizable text enlargement without changing the page layout. These updates are available in select English-speaking countries for devices running Android 15 and above.

May 15, 2025: Palantir, Legion Boost AI for US Spec Ops - Palantir Technologies and Legion Intelligence have partnered to integrate Generative AI (GenAI) into the U.S. Special Operations Command's Mission Command System. This collaboration enhances SOCOM's situational awareness, decision-making, and operational workflows by deploying GenAI across sensitive networks, leveraging Palantir's Apollo Mission Manager for secure AI deployment and superior mission outcomes.

Legion's platform enables AI incorporation into diverse applications, boosting SOCOM's operational technology capabilities. CEO Ben Van Roo emphasized GenAI's critical role in defense processes, highlighting the partnership's potential to empower warfighters with actionable intelligence globally.

May 15, 2025: Databricks $1B Neon Acquisition Boosts AI Strategies - Databricks has acquired serverless PostgreSQL startup Neon for $1 billion, emphasizing the importance of agile data solutions for AI development. Neon's technology, which separates storage and compute, supports agentic AI applications through automated scaling and rapid database deployment. This acquisition highlights the changing infrastructure needs for successful AI, suggesting that traditional databases can hinder AI progress.

Enterprises are urged to prioritize serverless capabilities to enable dynamic resource allocation, making database strategies more competitive for efficient AI solutions. The move by Databricks illustrates a growing trend towards serverless technology to enhance AI projects, pointing to a future where serverless databases will play a crucial role in AI development.

May 15, 2025: H2O.ai Achieves Key FedRAMP Milestone - H2O.ais AI Cloud has reached the FedRAMP In Process status at the High Impact Level, marking progress in delivering secure AI solutions for the U.S. public sector. This status allows H2O.ai to support sensitive government workloads, ensuring high-level security and compliance.

The company is committed to providing secure AI that federal agencies can trust, with applications in fraud detection and cybersecurity. This achievement reflects H2O.ais dedication to facilitating Sovereign AI in the public sector.

May 15, 2025: Meta Delays Llama 4 Behemoth Model Release - Meta has delayed the release of its Llama 4 Behemoth model from early summer to fall or later due to performance concerns. This postponement raises doubts about the effectiveness of Meta's significant AI investments and strategy, impacting its stock price. Internal issues and changes in research teams have contributed to the delay.

Other AI companies, like OpenAI and Anthropic, face similar setbacks, indicating a slowdown in AI advancements and increased costs industry-wide. As frustrations rise, possible management changes are anticipated within Meta's AI division.

May 15, 2025: AI Fuels Major Programmer Layoffs at Microsoft - Microsoft recently laid off 2,000 employees in Washington, with over 40% being software engineers, the largest impacted group. These layoffs are part of a larger reduction affecting 6,000 staff members globally. The job cuts aim to streamline management structures, the company states.

While CEO Satya Nadella noted that AI now writes up to 30% of Microsoft's code, the company has not confirmed if this directly links to the layoffs. Microsoft avoided further comment on the role of AI in these changes, focusing instead on restructuring efforts.

May 15, 2025: AI Citation Error Sparks Legal Embarrassment - An Anthropics lawyer apologized for using an incorrect citation from its Claude AI chatbot in a legal case with music publishers. The citation included wrong titles and authors, which were not caught during a manual check. Opposing lawyers pointed out the error, leading a federal judge to demand a response.

This incident is part of a broader trend of AI-related courtroom issues. However, it hasn't deterred investment in legal AI startups. For instance, Harvey, a legal AI company, is reportedly seeking substantial funding.

May 14, 2025: Googles Gemini Boosts GitHub Code Analysis - Google's Gemini chatbot now offers direct GitHub integration for Advanced plan subscribers at $20 per month, allowing users to import codebases for code generation, explanation, and debugging. This integration enhances functionality but AI models like Gemini still face challenges in producing high-quality, secure code.

This update parallels OpenAI's recent GitHub connector launch for ChatGPT deep research, marking rapid advancements in AI capabilities. Companies continue to strive for differentiation in a competitive market by enhancing their tools.