Desk276 storiesPage 12 of 12

AI.

Artificial intelligence, LLMs, and ML research

All desksAIDeFi & CryptoFintechHardwareSoftwareDrones & RobotsGaming
Editorial cover for Collibra's AI Command Center says enterprise AI is moving from model governance to live agent control
AI

Collibra's AI Command Center says enterprise AI is moving from model governance to live agent control

Collibra's May 6, 2026 launch of AI Command Center matters because it treats agentic AI as an operations problem, not just a model problem. As AI systems start taking actions across enterprise workflows, the winning control layer may be the one that can watch live behavior, enforce policy continuously, and step in before an agent mistake becomes a business incident.

5 min to readMay 7, 2026 · 05h
Editorial cover for UiPath's public-sector release says agentic AI is moving on-prem before many governments trust the cloud
AI

UiPath's public-sector release says agentic AI is moving on-prem before many governments trust the cloud

UiPath's May 5 and May 6 rollout of on-premises agentic AI capabilities for government buyers matters because it pushes the agent story into one of the most constrained enterprise environments. If public-sector teams want automation and agents without giving up data residency, air-gapped controls, and procurement discipline, the next AI battleground becomes deployment architecture rather than model novelty.

5 min to readMay 6, 2026 · 17h
Editorial cover for OpenAI and PwC's new CFO push says agentic AI is moving from copilots into finance operations
AI

OpenAI and PwC's new CFO push says agentic AI is moving from copilots into finance operations

OpenAI's May 4 collaboration with PwC and PwC's May 5 expansion notes point to a more consequential enterprise AI shift than another generic assistant rollout. The real signal is that finance, one of the most controlled and judgment-heavy functions inside large companies, is being used as a proving ground for human-governed agentic workflows across procurement, treasury, reporting, tax, and the close.

5 min to readMay 6, 2026 · 05h
Editorial cover for Washington's new CAISI deals turn frontier AI testing into pre-release infrastructure
AI

Washington's new CAISI deals turn frontier AI testing into pre-release infrastructure

The May 5 CAISI agreements matter because they shift frontier-model evaluation from an ad hoc safety ritual into something closer to critical infrastructure. When Microsoft, Google DeepMind, and xAI agree to let government testers examine unreleased systems, the AI race stops being only about launch speed and starts becoming a contest over who can prove operational trust before deployment.

5 min to readMay 5, 2026 · 17h
IBM logo used as visual context for the Think 2026 announcement.
AI

IBM's Think 2026 launch says the AI race is shifting from models to operating systems

IBM's May 5, 2026 Think announcements matter because they frame enterprise AI as an operating-model problem, not a demo problem. The center of gravity is moving toward agent orchestration, real-time data plumbing, governance, and sovereign control for workloads that have to survive audits, outages, and board-level scrutiny.

5 min to readMay 5, 2026 · 10h
OpenAI logo used as article visual context.
AI

OpenAI's Microsoft rewrite turns frontier AI into a multi-cloud market

OpenAI's late-April rewrite of its Microsoft relationship matters less as partnership drama than as market structure. By converting exclusivity into first-refusal economics, OpenAI widened its infrastructure options, protected Stargate-scale expansion, and signaled that frontier AI capacity is now too large to sit inside a single-cloud dependency.

4 min to readMay 2, 2026 · 17h
Illustrated cover for OpenAI's Advanced Account Security launch.
AI

OpenAI's new security tier shows AI accounts are becoming critical infrastructure identities

OpenAI's April 30 launch of Advanced Account Security is a small product change with a larger implication: frontier AI accounts are no longer casual logins. They are becoming high-value operational identities that can expose code, business context, and security-sensitive workflows if they are taken over.

4 min to readMay 2, 2026 · 05h
Illustrated cover for OpenAI's April 2026 compute infrastructure expansion and Microsoft partnership update.
AI

OpenAI's new compute push shows the AI race is turning into an infrastructure contest

OpenAI's April 29 infrastructure update and Microsoft's April 27 partnership reset point to the same strategic reality: frontier AI competition is no longer just about better models, but about who can finance, secure, and operate compute at multi-gigawatt scale.

4 min to readMay 1, 2026 · 05h
Google Cloud Gemini Enterprise artwork used to illustrate enterprise AI infrastructure.
AI

Big Tech's $700 billion AI spend is becoming a cloud revenue sorting machine

Alphabet's April 30 cloud breakout, Amazon's AWS results, and OpenAI's new Stargate update all point to the same shift: AI competition is no longer about demos alone, but about who can turn compute into durable revenue fastest.

5 min to readApr 30, 2026 · 17h
A DeepSeek smartphone image used in AP coverage of the V4 launch.
AI

DeepSeek V4 is turning Huawei's AI stack into a demand signal, not a workaround

DeepSeek's V4 launch has moved China's AI conversation from model hype to stack validation, with Huawei chips, cloud rollouts, and agent tooling all tightening into one commercial proof point.

4 min to readApr 30, 2026 · 09h
Abstract TechPulse cover for AI agents and compute economics.
AI

GPT-5.5 turns the AI race toward agents and compute economics

The next frontier-model competition is less about a single benchmark and more about agent reliability, tool use, memory, latency, and the cost of running useful work.

1 min to readApr 29, 2026 · 23h
Illustrative image used by Live Science for Claude Mythos coverage, representing restricted cyber-capable AI models.
AI

Anthropic Mythos turns frontier AI into a cybersecurity governance problem

Claude Mythos Preview is not just a stronger model. Its restricted rollout shows that frontier AI capability is becoming a cybersecurity access-control problem.

4 min to readApr 29, 2026 · 20h