Chronological index
Browse 1,687 source-linked AI reports by month.
Anthropic launched Claude Science in beta with more than 60 scientific skills and connectors, local and remote compute, SSH and HPC support, auditable artifacts, and reviewer-agent workflows.
Why it matters: Scientific assistants are becoming execution workbenches that connect literature, code, compute, and review rather than only answering questions.
Anthropic kept Claude Sonnet 5 at $2 per million input tokens and $10 per million output tokens instead of applying the previously scheduled 1 September increase to $3 and $15. The lower launch price is now the standard price.
Why it matters: A temporary workhorse-model discount became a durable price. That changes routing and agent-cost assumptions for teams that had budgeted for the announced increase.
Google released ADK Go 2.0 with a graph-based workflow engine, persistent state, dynamic orchestration, and human-in-the-loop pause and resume controls.
Why it matters: Agent frameworks are adding durable, deterministic workflow primitives instead of leaving every transition to model judgment.
Google moved Gemini Omni 1.1 Flash from developer preview to a stable Gemini API model. The asynchronous Interactions API now supports text, image, and video input, conversational video generation and editing, first-and-last-frame interpolation, three-to-ten-second output at 24 fps, and 360p, 720p, 1080p, or 4K controls; Google says 1080p and 4K output is upscaled.
Why it matters: Video generation and iterative editing now have a stable API identity, but resolution provenance, asynchronous job handling, moderation, rights, regional access, cost, and output consistency remain deployment decisions.
Google released Nano Banana 2 Lite, model ID gemini-3.1-flash-lite-image, for low-cost image generation and editing at a listed $0.034 per 1,000-pixel image.
Why it matters: Image generation is becoming an inexpensive production primitive suitable for high-volume variants, not only a premium creative feature.
IBM Research released a benchmark for multi-file Java modernization tasks where agents must migrate frameworks while preserving build behavior and passing repository tests.
Why it matters: Legacy modernization is a high-value coding-agent use case that requires dependency reasoning, coordinated edits and deterministic validation beyond isolated bug fixes.
Microsoft Research introduced a method that optimizes reusable agent-skill instructions against task feedback, treating the skill library as an explicit component to train and evaluate.
Why it matters: Skills can improve without changing the base model, but only if their versions, training tasks and held-out effects remain inspectable.
OpenAI released a 129-problem computational-biology benchmark across ten domains, built from synthetic data with known causal structure. Its best reported GPT-5.6 Sol result was 31.5% with Pro mode.
Why it matters: It targets the weak part of scientific agents: selecting analyses, revising assumptions, and closing the inferential loop under messy evidence.
Meta described a second-generation research system that translates non-invasive brain recordings into sentences in real time and released code plus research data for parts of the project.
Why it matters: The work suggests language decoding can be paired with non-invasive sensors and real-time models, a potentially important direction for assistive communication.
OpenAI reported that non-developer Codex users grew 137-fold among individuals and 189-fold in organizations since August 2025, while its heaviest internal users generated more than 60 agent-hours per day through parallel work.
Why it matters: Agent adoption is moving from coding into legal, recruiting, finance, analysis, and other knowledge work, with parallel agents changing the unit of work.
Google added browser and interface control as a built-in tool for Gemini 3.5 Flash through the Gemini API and its enterprise agent platform, with optional confirmation steps and indirect prompt-injection protections.
Why it matters: Computer control is becoming a standard model primitive rather than a bespoke agent wrapper.
Qualcomm unveiled its Dragonfly data-centre roadmap, including C1000 CPUs, AI300 accelerators, and high-bandwidth cache, and announced a multi-generation agreement with Meta.
Why it matters: A major mobile-chip supplier is making a full-stack attempt to enter AI data centres, adding another potential alternative to incumbent accelerator platforms.
Researchers introduced a red-team framework for training controls on difficult-to-grade research tasks. In experiments, a stronger model generated poor experiment proposals that a weaker scorer rated highly. Better scorer prompts improved robustness under the tested conditions.
Why it matters: Slow, repeated research degradation can evade oversight even when every proposal receives a favorable score. The framework tests that failure directly.
GitHub added BYOK support for OpenAI, Azure, Microsoft Foundry, Anthropic, LM Studio, Ollama, and OpenAI-compatible endpoints, storing local secrets in the operating system keychain.
Why it matters: Developer-agent interfaces are separating from model procurement, enabling private, local, and enterprise-controlled inference behind one workflow.
Mistral released a new OCR and document-intelligence model for extracting text, tables, equations, layout and image content through its API and platform.
Why it matters: Document pipelines need page structure and visual evidence, not just plain text, especially for regulated forms, scientific papers and complex tables.
Leaders of the Five Eyes cybersecurity agencies warned that frontier models could rapidly transform offensive and defensive cyber capabilities, saying the relevant timeline is months rather than years.
Why it matters: The warning places patching, identity, resilience, and AI-enabled defense on an immediate operating timeline rather than a long-term policy horizon.
Cloudflare introduced programmatically created, time-limited accounts for background agents to deploy and test resources without a human dashboard signup or permanent API token.
Why it matters: Ephemeral accounts reduce the blast radius and cleanup burden of agents that need real cloud infrastructure for a bounded task.
Adobe opened AI Assistant public betas in Premiere, Photoshop, Illustrator, InDesign, and Frame.io for multi-step production work such as organizing footage, batch edits, preflight checks, and feedback handling.
Why it matters: Agentic editing moved inside professional source files and timelines, where actions are more useful but can also change many assets or project states at once.
In an NEJM AI study, researchers used OpenAI o3 Deep Research to review 376 previously unsolved cases; expert review, further testing, and clinical confirmation established 18 diagnoses, a 4.8% additional yield.
Why it matters: This is stronger evidence than a medical benchmark because model-generated leads were filtered through established clinical processes and confirmed.
Hugging Face released a practical method for evaluating open models against an organization's actual tool schemas and tasks instead of relying only on generic agent benchmarks.
Why it matters: Tool names, schemas and error behavior materially affect model performance, so generic leaderboard scores can mislead a deployment decision.
Google published the Apache-2.0 Agentic Resource Discovery specification around an ai-catalog.json file and cryptographic verification, while GitHub shipped Agent Finder support for Copilot.
Why it matters: Agents need a scalable way to discover capabilities and verify what they are connecting to without hard-coded integration lists.
GitHub made the Copilot app generally available for assigning issues and coding tasks to cloud agents through repository workflows and reviewable pull requests.
Why it matters: Issue-driven delegation turns the tracker into an execution surface, making repository permissions, branch protection and review policy part of the agent contract.
Zhipu released GLM-5.2 weights and technical material focused on sustained coding, research and tool-use workloads with longer execution horizons.
Why it matters: Open long-horizon models give teams another controllable backend for testing whether persistence comes from weights, context management or harness design.
The European Commission's 2026 Digital Decade report says nearly 20% of EU enterprises use AI, up 48% year over year, while cloud and data-analytics adoption reached 46.7% and 39.9% respectively.
Why it matters: AI adoption is growing quickly but remains far from universal, and the report connects uptake to cloud, data, skills, and infrastructure constraints.
OpenAI described Deployment Simulation, a privacy-preserving method that regenerates responses to de-identified historical conversation prefixes to estimate how often undesirable behavior may appear after launch.
Why it matters: Traditional red teams find severe edge cases but say little about incidence. Simulation connects pre-release testing to the distribution of actual use.
Google DeepMind, Google Cloud, Faculty, and UK public-sector partners announced a prototype intended to help planning officers review applications, with a stated goal of cutting processing time by 50%.
Why it matters: This is a concrete attempt to apply AI to a document-heavy public workflow with an auditable human decision at the end.
Anthropic retired `claude-sonnet-4-20250514` and `claude-opus-4-20250514` from its first-party Claude API. Requests to either exact model ID now fail, requiring affected production workloads to move to supported successors and revalidate behavior.
Why it matters: Both IDs can remain embedded in gateways, jobs, fixtures, and fallback chains, creating hard failures at a first-party API boundary.
The OSGuard preprint introduces action-level and live OSWorld evaluations containing latent hazards, reporting that guardrails which look strong in isolation can still fail when embedded in complete computer-use trajectories.
Why it matters: Agent safety must be evaluated at the action and trajectory level, where harmless-looking individual steps can combine into a dangerous outcome.
Anthropic says a U.S. export-control action on 12 June caused it to suspend Fable 5 globally; approvals for controlled Mythos 5 deployments followed on 26 June, controls were lifted on 30 June, and Fable returned on 1 July.
Why it matters: Frontier-model availability can change abruptly through export policy, even for customers far from a sanctioned end use.
AWS released Agent-EvalKit for defining datasets, evaluators, experiments and reports across agent systems, with support for deterministic and model-based checks.
Why it matters: A portable evaluation harness helps teams compare framework or model changes without rebuilding measurement logic for every agent stack.
A June preprint re-evaluated enterprise agents on WorkBench and reported 98% task success for Claude Fable 5 versus 43% for GPT-4, while harmful-action rates fell from 26% to 1.9%.
Why it matters: Capability and safety can improve together, but a low percentage of harmful actions is still material when agents operate at enterprise scale.
A NIST scientist published a formal argument, based on incompleteness logic, that a finite fixed set of guardrails cannot be universally robust against adaptive adversarial prompts. NIST recommends a continuous monitor-and-update security model rather than one-time certification.
Why it matters: Guardrail deployment is an operating process: attackers adapt, contexts change and a static ruleset cannot be treated as a permanent safety proof.
Anthropic released Claude Fable 5 for advanced agentic and coding work at $10 per million input tokens and $50 per million output tokens, while limiting the more cyber-capable Mythos 5 to trusted deployments.
Why it matters: The split release makes capability-tiered access an explicit product strategy: broad access for a powerful work model and controlled access for higher-risk cyber capability.
Cohere Labs released North Mini Code, its first developer-focused open model, with weights and evaluation material intended for efficient coding and agentic tasks.
Why it matters: A compact coding model can handle routine local work cheaply and privately if routing and deterministic checks catch tasks that require stronger models.
Google introduced continuous speech-to-speech translation across more than 70 languages in Gemini Live, with public API and AI Studio preview access plus narrower previews in Meet and Google Translate.
Why it matters: Real-time translation is moving from segmented transcription pipelines toward a single conversational audio system that preserves timing and interaction.
Google DeepMind reported a pre-registered eight-week trial involving 1,763 students across 12 schools, with the guided-learning group improving mathematics scores by 0.258 standard deviations.
Why it matters: The study measures learning outcomes rather than usage or satisfaction, which is the right standard for educational AI.
AWS published a test harness that feeds synthetic audio into Amazon Nova Sonic sessions and captures transcripts, timing, tool behavior and outcomes for repeatable automated evaluation.
Why it matters: Voice agents need scalable tests for interruption, latency, accents and tool execution that do not depend on a human repeatedly speaking into a microphone.
Microsoft open-sourced 100 time-evolving tasks across ten synthetic web apps, including no-op cases and conditions where agents should watch, wait and act only after change.
Why it matters: Always-on agents waste money and create risk when they poll or act continuously; waiting strategy can matter as much as model choice.
Google upgraded NotebookLM with Gemini 3.5, an isolated cloud computer, more than 100 agentic skills, and structured outputs including charts, spreadsheets, and slides.
Why it matters: Research assistants are evolving from source-grounded chat into artifact-producing agents with execution environments.
SK Telecom and NVIDIA announced plans for a gigawatt-scale AI cloud built around NVIDIA DSX infrastructure, with the first AI-factory phase targeted for 2027.
Why it matters: National and regional AI capacity is being assembled as complete cloud and factory infrastructure, not simply individual GPU purchases.
OpenAI's compliance guidance marks 5 June as the removal date for the older stateful conversation-log route deprecated when its newer conversation-logs system launched on 5 March. Enterprise and Edu integrations still using the legacy route require migration.
Why it matters: A stale compliance integration can fail silently or interrupt SIEM, eDiscovery, data-loss prevention, investigation, and retention workflows that depend on complete logs.
OpenAI began rolling out a new Dreaming memory architecture that synthesizes context across conversations, updates time-sensitive facts and exposes a reviewable memory summary. It launched first for US Plus and Pro users, with broader and lower-tier rollout planned.
Why it matters: Long-lived assistants need memory that can revise stale context and show users what was retained, not only accumulate explicit notes.
Anthropic analyzed 832 accounts banned for malicious cyber activity from March 2025 to March 2026. It found 67.3% used AI for malware preparation, 6.5% for lateral movement, and a shift toward later attack stages; it also argues MITRE ATT&CK lacks AI-specific activity labels.
Why it matters: Defenders should model AI as a capability that can connect multiple attack stages, not only as a phishing-copy accelerator.
Google released Gemma 4 12B weights under Apache 2.0 with text, vision, and native audio understanding, and says the model can run in roughly 16 GB of memory.
Why it matters: A compact, permissively licensed multimodal model expands practical local inference beyond text and images into audio.
Microsoft reported that Infosys, TCS, and Wipro had each deployed more than 100,000 Microsoft 365 Copilot licences, exceeding 300,000 employees in total.
Why it matters: India's largest IT-services companies are moving enterprise copilots from pilots to workforce-scale deployments.
Microsoft announced Agent Harness, hosted execution, CodeAct-style tool use and related framework updates for building and operating agents across its development stack.
Why it matters: Separating the harness from the model makes execution policy, state, tools and hosting explicit components that teams can version and evaluate.
The MPBench preprint identifies four memory-write channels, nine memory-system vulnerabilities, and six attack classes, finding that agents which write memory more aggressively can become easier to manipulate.
Why it matters: Persistent memory creates a durable attack surface: one malicious interaction can influence later tasks even after the original content has disappeared.
OpenAI added GPT-6 Rosalind Research to Work, Codex and the API for a limited set of approved Research customers. The API model ID is gpt-6-rosalind; it is not in Chat, Discovery approval is insufficient, and Enterprise begins with a two-week admin preview that is off by default.
Why it matters: Life-sciences teams can now route approved multistep research to GPT-6 reasoning, while the separate Research approval and disabled-by-default enterprise rollout make authorization and model selection concrete operating controls.
A UN University report projected data-centre electricity use could reach 945 TWh, water consumption 9.3 trillion litres, and land use 14,500 square kilometres by 2030 under its modeled scenarios.
Why it matters: AI infrastructure decisions shift costs into electricity systems, water basins, land, and local communities that token pricing does not reveal.
GitHub launched Agent Apps that can be installed from its marketplace and invoked through issue assignment, mentions, and native repository interfaces, with an initial group of nine partners.
Why it matters: Agent distribution is becoming a platform layer: agents can enter existing issue and review workflows instead of requiring separate dashboards.
GitHub made its Copilot SDK generally available as a stable agent-runtime API for planning, tool use, file operations, streaming, and resumable sessions across six programming languages.
Why it matters: A mainstream coding platform is exposing the same runtime primitives behind its agent as an embeddable developer platform.
At Build, Microsoft introduced or expanded seven MAI models, including the 35B-active-parameter MAI-Thinking-1 and new image, transcription, voice, and low-latency coding models. Several media models reached Foundry while the reasoning model stayed with select early partners.
Why it matters: Microsoft is moving from being primarily a distributor of partner models toward owning a vertically integrated portfolio for distinct production workloads.
Microsoft previewed Scout, an always-on agent built on OpenClaw with Entra identity and organizational controls, alongside an alpha OpenClaw-on-Windows environment using OpenShell sandboxing.
Why it matters: Persistent agents create an identity and governance problem as much as a model problem; Microsoft is making enterprise identity part of the agent architecture.
A U.S. executive order directed agencies to pursue voluntary agreements under which participating developers would provide frontier models for cybersecurity testing up to 30 days before release.
Why it matters: Government model testing is shifting closer to release time, but a voluntary regime may produce uneven coverage across labs.
AWS organized production agent operations around governance and security, build and operations, evaluation and observability, with versioned artifacts, multi-account controls and release gates.
Why it matters: Agent operations span identity, tools, memory, traces, evaluation, cost and incident response; treating them as ordinary prompt deployments leaves critical gaps.
GitHub moved all Copilot plans to AI Credits, with included monthly allocations and paid overage governed by budgets. Copilot code review also began consuming GitHub Actions minutes, while user-level budget controls became generally available.
Why it matters: Agent and review activity now crosses multiple metered resources, changing cost allocation, budget enforcement, procurement assumptions, and operational controls for every Copilot plan.
Google reached the shutdown date for `gemini-2.0-flash`, `gemini-2.0-flash-001`, `gemini-2.0-flash-lite`, and `gemini-2.0-flash-lite-001` in the Gemini API. Applications pinned to any of the four IDs require migration to a supported model.
Why it matters: Undiscovered endpoint references can fail in production, while a hurried successor swap can change schemas, grounding, latency, safety, quotas, and cost.
76 stories
Browse month November 2026324 stories
Browse month October 2026296 stories
Browse month September 2026281 stories
Browse month August 2026144 stories
Browse month July 202657 stories
Browse month June 202649 stories
Browse month May 202646 stories
Browse month April 202645 stories
Browse month March 202644 stories
Browse month February 202640 stories
Browse month January 202626 stories
Browse month December 202519 stories
Browse month November 202521 stories
Browse month October 202523 stories
Browse month September 202520 stories
Browse month August 202525 stories
Browse month July 202531 stories
Browse month June 202528 stories
Browse month May 202525 stories
Browse month April 202521 stories
Browse month March 202522 stories
Browse month February 202524 stories
Browse month January 2025