Nicheloom

The opportunity tracker for new startups.

AI News

Real headlines from AI-dedicated outlets (TechCrunch AI, The Verge AI, OpenAI News, Hugging Face, and more) plus AI-relevant Hacker News stories, pulled every 6 hours — not another analysis, just the fastest real feed of what's happening.

The Sources/Vertical/Function filters above don't apply here — those filter launches, and a news headline isn't a launch. Use Source below to narrow by news outlet instead.

Source

Headline volume, last 14 days

Showing 7001–7050 of 9246 headlines.

IBM asks DBAs to trust AI to act on their behalf May 5, 2026, 5:29 p.m.

The Register AI/ML

IBM has introduced new AI-powered automation features to its Db2 database platform, developed with assistance from Google and Intel. The update encourages database administrators to delegate routine operational tasks to autonomous AI agents.

ServiceNow clears agents for landing with new AI control tower May 5, 2026, 5 p.m.

The Register AI/ML

ServiceNow has launched a new AI control tower to oversee autonomous agents and enterprise AI workflows. The monitoring solution integrates technology from the company's recent acquisitions of Veza and Traceloop.

GPT-5.5 Instant: smarter, clearer, and more personalized May 5, 2026, 10 a.m.

OpenAI News

OpenAI has introduced GPT-5.5 Instant as the updated default model for ChatGPT. The release provides more accurate answers with fewer hallucinations, alongside enhanced controls for personalization.

New ways to buy ChatGPT ads May 5, 2026, midnight

OpenAI News

OpenAI is broadening commercial access to ChatGPT ads by launching a beta self-serve Ads Manager, cost-per-click bidding, and improved measurement features. The advertising tools are designed to maintain user privacy while keeping sponsored promotions separate from conversations.

Advancing youth safety and wellbeing in EMEA May 5, 2026, midnight

OpenAI News

OpenAI has announced the European Youth Safety Blueprint along with EMEA Youth & Wellbeing Grants. The initiatives are aimed at fostering safe and responsible AI usage for teenagers, families, and educators across the region.

Microsoft fixes VS Code after app gives Copilot credit for human's work May 4, 2026, 9:13 p.m.

The Register AI/ML

Microsoft has issued a fix for VS Code after its Git extension began automatically listing GitHub Copilot as a co-author on commits. The default setting drew complaints from developers who objected to the AI tool taking credit for human work.

OpenAI and PwC collaborate to reimagine the office of the CFO May 4, 2026, 9 p.m.

OpenAI News

OpenAI and PwC have formed a partnership to help enterprise organizations modernize the office of the CFO using AI agents. The collaboration aims to automate finance workflows, enhance forecasting accuracy, and reinforce internal financial controls.

AI inference just plays by different rules May 4, 2026, 3 p.m.

The Register AI/ML

Existing cloud storage architectures are facing new strains because they were not built to meet the data demands of agentic AI. As AI inference workloads evolve, infrastructure must shift to handle operational requirements that differ sharply from traditional computing.

How OpenAI delivers low-latency voice AI at scale May 4, 2026, midnight

OpenAI News

OpenAI has rebuilt its WebRTC stack to power real-time voice AI with low latency on a global scale. The upgraded infrastructure is engineered to facilitate natural conversational turn-taking during audio interactions.

DeepSeek begins limited testing of vision mode April 30, 2026, 7:19 a.m.

TechNode (China)

DeepSeek has launched limited testing of a new vision mode across its web and app interfaces. Positioned alongside Fast Mode and Expert Mode, the update represents the platform's initial expansion into multimodal capabilities beyond standard OCR text recognition.

Introducing Advanced Account Security April 30, 2026, midnight

OpenAI News

Advanced Account Security has been introduced to help prevent account takeovers and protect sensitive user data. The updates include phishing-resistant logins, improved recovery methods, and enhanced security protections.

Where the goblins came from April 29, 2026, 8 p.m.

OpenAI News

A new post details how personality-driven 'goblin' outputs originated and spread within GPT-5. It covers the timeline, root causes, and fixes implemented to address the behavioral quirks.

Cybersecurity in the Intelligence Age April 29, 2026, 4 a.m.

OpenAI News

OpenAI has published a five-part action plan aimed at strengthening cybersecurity in the Intelligence Age. The strategy centers on democratizing AI-driven cyber defense and safeguarding critical digital systems.

TSMC accelerates 2nm expansion, targets record five-fab ramp in 2026 April 29, 2026, 2:32 a.m.

TechNode (China)

TSMC is doubling its advanced-node expansion pace to address surging demand for AI and high-performance computing. According to Senior Vice President Hou Yongqing, five 2nm fabrication facilities are slated to begin ramping up for mass production in 2026.

DeepSeek V4 report shows multiple R&D staff departures April 28, 2026, 3:22 a.m.

TechNode (China)

The technical report for DeepSeek V4 revealed that 10 contributors out of its nearly 300-person author list have left the company. Departures reportedly include at least five core R&D members since late 2025 across key areas such as base models, reasoning, OCR, and multimodality.

Our commitment to community safety April 28, 2026, midnight

OpenAI News

OpenAI has detailed its approach to maintaining community safety across ChatGPT. The framework relies on model safeguards, misuse detection systems, policy enforcement, and ongoing work with safety experts.

OpenAI models, Codex, and Managed Agents come to AWS April 28, 2026, midnight

OpenAI News

OpenAI's GPT models, Codex, and Managed Agents are now accessible on AWS. The integration allows enterprise customers to build secure AI applications directly within their existing AWS environments.

OpenAI available at FedRAMP Moderate April 27, 2026, 2 p.m.

OpenAI News

ChatGPT Enterprise and the OpenAI API have achieved FedRAMP Moderate authorization. The milestone allows U.S. federal agencies to securely adopt and deploy OpenAI's technologies.

The next phase of the Microsoft OpenAI partnership April 27, 2026, 6 a.m.

OpenAI News

OpenAI and Microsoft have announced an amended partnership agreement. The update is designed to simplify their relationship, provide long-term clarity, and support ongoing AI innovation at scale.

Intel warns China of severe server CPU shortage as AI demand surges April 27, 2026, 2:53 a.m.

TechNode (China)

Intel has warned major cloud service providers in China of severe server CPU shortages across the second and third quarters. The constraints stem from surging AI infrastructure demand and slower-than-expected yield ramp-up for its 18A process, leading to extended delivery lead times.

DeepSeek V4 becomes default model for OpenClaw April 27, 2026, 1:48 a.m.

TechNode (China)

Open source agent framework OpenClaw has integrated DeepSeek V4 models into its core ecosystem with its 2026.4.24 release. DeepSeek V4 Flash has been set as the default model, offering performance near that of the 1.6-trillion-parameter V4 Pro while utilizing 284 billion total and 13 billion activated parameters.

Choco automates food distribution with AI agents April 27, 2026, midnight

OpenAI News

Food distribution platform Choco leveraged OpenAI APIs to automate its operations and improve productivity. A customer story highlights how the company implemented AI agents to streamline distribution and drive business growth.

An open-source spec for orchestration: Symphony April 27, 2026, midnight

OpenAI News

Symphony is an open-source specification designed for Codex orchestration. It converts issue trackers into always-on agent systems to increase engineering output and reduce context switching.

Our principles April 26, 2026, 4 p.m.

OpenAI News

OpenAI CEO Sam Altman has shared five core principles guiding the organization's work. The principles align with the company's stated mission to ensure that artificial general intelligence benefits all of humanity.

Modeling an AI jobs transition April 25, 2026, midnight

OpenAI News

A new analytical framework evaluates 921 occupations covering 148 million U.S. jobs to assess the impact of artificial intelligence. The model categorizes roles based on whether they face automation risk, reorganization, growth, or minimal disruption.

DeepSeek-V4 preview now available with open-source access April 24, 2026, 7:23 a.m.

TechNode (China)

DeepSeek has released an open-source preview of its V4 model series, featuring an ultra-long context window of up to one million tokens. Available in Pro and Flash versions, the models offer upgraded reasoning, knowledge coverage, and agent capabilities.

Introducing GPT-5.5 April 23, 2026, 11 a.m.

OpenAI News

OpenAI has introduced GPT-5.5, a model designed for complex workloads across diverse tools. It focuses on capabilities such as coding, research, and data analysis.

Everyday work with ChatGPT Work April 23, 2026, 10 a.m.

OpenAI News

A new guide outlines practical everyday workflows using ChatGPT Work. It details how teams can apply the tool to recurring tasks, research, updates, planning, and general operations.

Making ChatGPT better for clinicians April 22, 2026, 3 p.m.

OpenAI News

OpenAI has made ChatGPT for Clinicians available for free to verified U.S. physicians, nurse practitioners, and pharmacists. The offering is designed to assist medical professionals with clinical care, medical documentation, and research.

Workspace agents April 22, 2026, 10 a.m.

OpenAI News

ChatGPT has introduced guides on creating and deploying workspace agents. These agents allow users to connect tools, automate repeatable workflows, and streamline routine team operations.

Introducing workspace agents in ChatGPT April 22, 2026, 10 a.m.

OpenAI News

ChatGPT has introduced workspace agents powered by Codex to automate complex workflows. Operating securely in the cloud, these agents assist teams in scaling tasks across multiple tools.