MIT Report: 95% of Enterprise GenAI Projects Fail to Deliver ROI
A new report from MIT’s NANDA initiative reveals that a staggering 95% of enterprise generative AI pilot programs are failing to produce a significant return on investment. Despite investments totaling between $35 and $40 billion in the US, only 5% of organizations have successfully deployed AI tools into production at scale. The report, titled “The GenAI Divide: State of AI in Business 2025,” is based on interviews with enterprise leaders, analysis of public AI initiatives, and a survey of business professionals. The primary reasons for failure are not a lack of infrastructure or talent, but rather the inability of AI systems to retain data, adapt, and learn over time. The study also found that purchasing AI tools from vendors and forming partnerships has a significantly higher success rate than in-house development.
Sources:
MIT Researchers Demystify Protein Language Models’ ‘Black Box’
Researchers at MIT have developed a new technique to understand the inner workings of protein language models, which have been instrumental in biological applications like drug target identification. These AI models have been largely treated as “black boxes,” with their decision-making processes unclear. The new method allows scientists to determine which protein features the models consider when making predictions. This insight can help researchers select the most appropriate models for specific tasks, thereby streamlining the development of new drugs and vaccines. The study, published in the Proceedings of the National Academy of Sciences, has broad implications for improving the explainability of AI in scientific research.
Sources:
CNCF Aims to Standardize AI/ML Workloads on Kubernetes
The Cloud Native Computing Foundation (CNCF) is developing a certification program for Kubernetes distributions to ensure they can effectively run various AI workloads. The goal is to establish a set of requirements and recommendations to enable the seamless migration of AI models and inferencing workloads across different cloud environments. This initiative aims to replicate the success of the CNCF’s existing conformance guide for Kubernetes. By creating a standardized environment for AI/ML tasks on Kubernetes, the CNCF intends to eliminate the complexities associated with custom configurations. The foundation is currently seeking input from the community to help define these crucial requirements.
Sources:
Robotics & Autonomous Vehicle News: Humanoid Robot Trends and Tesla Scrutiny
This week in robotics and autonomous vehicles, the International Federation of Robotics released a new paper on the trends and opportunities in humanoid robot adoption. In the autonomous vehicle sector, recent lawsuits are scrutinizing Tesla’s self-driving technology, while the overall market for autonomous vehicles is projected to see significant growth, with key players like Tesla and Waymo expanding their services.
Sources:
Oracle and Google Cloud Partner to Bring Gemini AI to OCI
Oracle and Google Cloud have broadened their partnership to provide customers with access to Google’s advanced AI models, beginning with Gemini 2.5, through the Oracle Cloud Infrastructure (OCI) Generative AI service. This integration will allow Oracle customers to utilize Gemini models for a variety of use cases, including multimodal understanding, software development, and workflow automation. Oracle intends to make the full range of Gemini models available via the OCI Generative AI service through new integrations with Vertex AI. The collaboration will also extend to making Gemini models available within Oracle Fusion Cloud Applications in the future.
Sources:
Xerox Unveils AI-Powered EveryDoc IDP App for Automated Workflows
Xerox has released the EveryDoc IDP App, a solution built on its Intelligent Document Processing platform. The application uses advanced AI models to automate the extraction and verification of data, aiming to streamline document workflows. It is designed for both large enterprises and small- to medium-sized businesses to reduce manual document processing. The goal of the app is to enable faster and more informed decision-making without requiring a significant IT overhaul.
Sources:
AWS Weekly Roundup: New EC2 P5 Instances, Go Driver, and EKS Cilium Support
Amazon Web Services (AWS) has announced the general availability of new Amazon EC2 P5 instances featuring a single NVIDIA H100 GPU, designed to provide a cost-effective solution for machine learning and high-performance computing workloads. Additionally, AWS has released the AWS Advanced Go Driver for Amazon RDS and Amazon Aurora, which offers faster failover times and enhanced authentication options. For Kubernetes users, AWS has expanded its support for Cilium with Amazon EKS Hybrid Nodes, enabling broader feature availability for networking capabilities in containerized environments. These updates were part of the AWS Weekly Roundup for August 18, 2025.
Sources:
Sentry Launches MCP Tool for Application Performance Monitoring
Sentry has officially launched its new MCP monitoring tool, designed to provide developers with deeper insights into their application’s performance. The tool aims to help engineering teams identify and resolve performance bottlenecks more efficiently. By offering detailed performance data, Sentry’s MCP tool assists in optimizing code and improving the end-user experience. This release is part of a broader trend of new AI-powered development tools.
Sources:
bitHuman Releases SDK for Developing Lifelike AI Avatars
bitHuman has released a new software development kit (SDK) that enables developers to create realistic AI-powered avatars. This SDK is designed to facilitate the integration of interactive, human-like avatars into a variety of applications, from customer service to virtual experiences. The launch of bitHuman’s SDK provides developers with the tools to build more engaging and personalized user interfaces. The release reflects the growing interest in AI-driven human-computer interaction.
Sources: