Find published research by company, product, platform or technology.
Showing 1–20 of 67 dossiers
AgentControl now spans more production stacks: applications can resolve different prompts and models by context, track token/cost behavior, require approvals, use Bedrock without proxying inference through LaunchDarkly, and inspect multi-step agent runs as one conversation.
The new RubyGems evidence reinforces the same systems lesson already visible across Hugging Face, DseWiki and at least 10 other sites: supposedly isolated agents can repurpose reachable internet infrastructure in ways their operators did not intend.
The change separates three things that are often bundled together: the harness, the subscription that pays for it, and the sandbox that executes it. Builders can switch among supported coding agents behind one interface while reusing existing subscription access and reducing credential exposure inside agent runtimes.
A third-party GEO dataset recorded an 86.4% relative collapse in Reddit’s visible ChatGPT Search citation share while Google AI citation changes were much smaller. The result is a useful warning against building an AI-discovery strategy around one source platform, not proof of an OpenAI penalty or Reddit removal.
The useful change is containment rather than another browser-agent feature. Teams can let an agent operate a real browser while constraining its HTTP and HTTPS reach to the site and dependencies the task actually needs, reducing the blast radius of prompt injection, bad tool decisions or untrusted page content.
The material change is that model routing is no longer a single opaque optimization target. Developers can now tell Copilot whether to bias Auto toward lower cost, a middle ground or higher quality while GitHub still chooses a model prompt by prompt.
The material change is not another Meta model launch. Muse packages persistent autonomous execution, credentials, payments, app access and memory into a mainstream consumer product, making permission design and agent containment part of ordinary personal software rather than an enterprise-only problem.
The pilot attacks a persistent evaluation trade-off: labs do not want to reveal frontier-model internals, while evaluators do not want benchmark prompts leaking back to the model provider. DeepMind says a Singapore AI Safety Institute pilot kept both sides’ sensitive assets hidden during execution.
Ahrefs is standardising Brand Radar on an estimated AI-demand metric because major AI platforms do not publish prompt volume. The new number can improve relative weighting between prompts and platforms, but it remains a modelled proxy rather than a count of how many people actually asked an AI system a question.
The important change is at the gateway boundary, not just inference placement. OpenRouter says prompts can now stay in-region from decryption through provider execution and supported server tools, while teams can enforce the rule per workspace, team or API key.
SwarmLLM does not route whole prompts to separate machines; it pipelines one model across browser tabs. A MacBook and iPhone can jointly hold Qwen 3.8 27B even when neither device can hold the full 15GB quantized model alone, with no inference server in the loop.
Google is changing Gemini Notebook’s packaging from feature-style quotas toward a compute budget. That gives users more flexibility but makes the effective cost of one request less predictable and ties premium upgrades more directly to computational intensity.
Private Safety Processing is OpenAI’s attempt to reconcile stronger multi-turn safety monitoring with Zero Data Retention. Early customers are testing it now, with rollout and a technical white paper planned for September; important implementation details remain unpublished.
Meta has made the privacy-versus-price trade explicit in its Model API: developers can choose standard pricing or a contributor model ID with steeply discounted inference in exchange for training-data permission. The choice matters for proprietary code, customer data and AI SaaS workloads.
WebMCP has crossed from a browser experiment into usable platform integration: ChatGPT’s built-in browser discovers site tools, Chrome exposes the proposed standard experimentally, and WordPress Playground now bridges plugin-defined tools from embedded WordPress into that agent-facing layer.
The Assistants API shutdown date has passed. OpenAI’s deprecation documentation lists August 26, 2026 as the removal date and directs developers to Responses and Conversations for replacement workloads.
Quattro’s unified programmable shell is a real architecture change rather than a theme refresh. Omarchy 4.0.2 now hardens package, installer, SSH and input paths, while current user reports of Quickshell crashes and a runaway-memory event illustrate the new central shell’s blast radius.
GitHub’s new token-type controls let responders revoke PATs, SSH keys, OAuth app tokens or GitHub App user tokens selectively instead of invalidating every credential belonging to a user.
The important development is not simply another AI security mishap. Anthropic found a fourth incident missed by its first review, widened the search to hundreds of millions of transcripts, revised its causal interpretation and invited an external evaluator to investigate the full record.
This is not one headline vulnerability fix. Gemini CLI 0.60 is a coordinated hardening pass across the plumbing that lets extensions, sandboxes, filesystem paths and MCP authentication influence an agent’s execution environment.