<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
  <title>Slash Digital Lab · Insights</title>
  <link>https://slash-digital.io/en/insights/</link>
  <description>Sourced analysis: models, local AI, image and video, agents, cybersecurity and regulation.</description>
  <language>en</language>
  <atom:link href="https://slash-digital.io/en/insights/feed.xml" rel="self" type="application/rss+xml"/>
  <lastBuildDate>Mon, 12 Oct 2026 17:00:00 GMT</lastBuildDate>
  <item>
    <title>Chatbot vs AI agent: differences and which one you need</title>
    <link>https://slash-digital.io/en/insights/chatbot-vs-ai-agent/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/chatbot-vs-ai-agent/</guid>
    <pubDate>Mon, 12 Oct 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Agents and engineering</category>
    <description>A chatbot answers; an AI agent decides its steps and acts in your systems. Differences, a decision table, costs, risks and when a plain workflow is enough.</description>
  </item>
  <item>
    <title>Grok Bot with Claude Opus 5.5: why we prefer it</title>
    <link>https://slash-digital.io/en/insights/grok-bot-claude-opus-5-5/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/grok-bot-claude-opus-5-5/</guid>
    <pubDate>Thu, 08 Oct 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Agents and engineering</category>
    <description>Why Slash prefers Grok Bot, now announced with Claude Opus 5.5, over Meta's Muse and OpenAI's Dots: models, teams, availability, data and caveats.</description>
  </item>
  <item>
    <title>Event networking app: how to build one that works</title>
    <link>https://slash-digital.io/en/insights/event-networking-app/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/event-networking-app/</guid>
    <pubDate>Thu, 08 Oct 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Strategy and regulation</category>
    <description>How to build an event networking app: invitation-only access, agenda, directory, 1:1 meetings, offline mode, AI and app store rules, with real examples.</description>
  </item>
  <item>
    <title>Claude Opus 5.5: release date, pricing and what changes</title>
    <link>https://slash-digital.io/en/insights/claude-opus-5-5/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/claude-opus-5-5/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Models</category>
    <description>Claude Opus 5.5 (September 22, 2026): $4/$20 per million tokens, Fable 5.1-level results per Anthropic, breaking changes and when to use it.</description>
  </item>
  <item>
    <title>Best AI models, October 2026: Opus 5.5, GPT-6, Gemini</title>
    <link>https://slash-digital.io/en/insights/ai-models-2026/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/ai-models-2026/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Models</category>
    <description>The AI model map for October 2026: Claude Opus 5.5 and Sonnet 5.5, GPT-6 Astra, Gemini 4 Argon and Grok 4.7, with prices, leaderboards and how to choose.</description>
  </item>
  <item>
    <title>Prompt injection and AI agent security | Slash Digital Lab</title>
    <link>https://slash-digital.io/en/insights/prompt-injection-agents/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/prompt-injection-agents/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Cybersecurity</category>
    <description>Prompt injection explained: why it is still unsolved, 2025–2026 incidents, OWASP and MITRE ATLAS, defenses and their limits, and a checklist for agents.</description>
  </item>
  <item>
    <title>Local AI hardware in 2026: what to buy, what runs</title>
    <link>https://slash-digital.io/en/insights/local-ai-hardware-2026/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/local-ai-hardware-2026/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Local AI and open weights</category>
    <description>Local AI hardware in 2026: RTX 5090, RTX PRO 6000, DGX Spark, Ryzen AI Max+ and Mac Studio M5, with memory math, what fits where and buy-vs-rent costs.</description>
  </item>
  <item>
    <title>Image models in 2026: which to use for brand work</title>
    <link>https://slash-digital.io/en/insights/image-models-2026/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/image-models-2026/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Image, video and audio</category>
    <description>Image models in September 2026: Nano Banana, GPT Image 2.5 and open models, price per image, brand-safe use and the EU AI Act Article 50 labeling rules.</description>
  </item>
  <item>
    <title>AI agents in production: what works in 2026</title>
    <link>https://slash-digital.io/en/insights/ai-agents-production/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/ai-agents-production/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Agents and engineering</category>
    <description>AI agents in production in 2026: workflow or agent, adoption and failure data, design rules, cost control, security and a launch checklist.</description>
  </item>
  <item>
    <title>AI Act, Colombia and ISO 42001: the 2026 AI regulatory map</title>
    <link>https://slash-digital.io/en/insights/ai-regulation-2026/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/ai-regulation-2026/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Strategy and regulation</category>
    <description>The 2026 AI regulatory map: AI Act dates after the Omnibus, GDPR, Colombia's Law 1581 and SIC rules, LatAm, the US, ISO 42001 and a 10-step plan.</description>
  </item>
  <item>
    <title>Uncensored AI models: what they are, uses and risks</title>
    <link>https://slash-digital.io/en/insights/uncensored-models/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/uncensored-models/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Local AI and open weights</category>
    <description>Uncensored AI models: what they are, which build fits 32 to 512 GB (October 2026), what removing safeguards breaks and what the law says.</description>
  </item>
  <item>
    <title>Video models in 2026: from demo to ad | Slash Digital Lab</title>
    <link>https://slash-digital.io/en/insights/video-models-2026/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/video-models-2026/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Image, video and audio</category>
    <description>AI video models in September 2026: Omni Flash, Veo 3.1, Kling, Seedance, price per second, the Sora shutdown, DOOH and AI Act labeling for brands.</description>
  </item>
  <item>
    <title>What is MCP (Model Context Protocol)? A guide for companies</title>
    <link>https://slash-digital.io/en/insights/mcp-explained/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/mcp-explained/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Agents and engineering</category>
    <description>What the Model Context Protocol (MCP) is, how it evolved to the 2026-07-28 spec, governance, real security risks and an adoption plan for companies.</description>
  </item>
  <item>
    <title>Deepfake and cloned-voice fraud: a protocol for companies</title>
    <link>https://slash-digital.io/en/insights/deepfake-fraud/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/deepfake-fraud/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Esteban Zuleta</dc:creator>
    <category>Cybersecurity</category>
    <description>Real cases, FBI figures and a verification protocol against deepfake and cloned-voice fraud for finance, HR and executive assistants.</description>
  </item>
  <item>
    <title>Open-weight models 2026: Kimi K3, Qwen3.8, DeepSeek V4.1</title>
    <link>https://slash-digital.io/en/insights/open-weight-models-2026/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/open-weight-models-2026/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Local AI and open weights</category>
    <description>Open-weight models as of September 2026: Kimi K3, Qwen3.8, DeepSeek V4.1, GLM-5.3, Gemma 4 and more, with sizes, licenses, benchmarks and risks.</description>
  </item>
  <item>
    <title>Reasoning effort in GPT, Claude and Gemini: when it pays</title>
    <link>https://slash-digital.io/en/insights/reasoning-effort/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/reasoning-effort/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Models</category>
    <description>How Claude, GPT-6 and Gemini bill thinking, where reasoning effort pays off, where it doesn't, and how to tune it by cost per solved task.</description>
  </item>
  <item>
    <title>Agentic coding in 2026: how a software team changes</title>
    <link>https://slash-digital.io/en/insights/agentic-coding/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/agentic-coding/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Agents and engineering</category>
    <description>Agentic coding in 2026: Claude Code, Codex, Copilot and Cursor, what METR and DORA measured, security risks and the team metrics that matter.</description>
  </item>
  <item>
    <title>How AI changes pentesting: attack and defense in 2026</title>
    <link>https://slash-digital.io/en/insights/ai-offensive-security/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/ai-offensive-security/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Esteban Zuleta</dc:creator>
    <category>Cybersecurity</category>
    <description>What attackers do with AI, what Mythos, Big Sleep and XBOW find, and how to adapt pentesting in Colombia: frequency, APIs, AI features, retests.</description>
  </item>
  <item>
    <title>Voice AI in 2026: TTS, transcription and voice agents</title>
    <link>https://slash-digital.io/en/insights/voice-audio-ai-2026/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/voice-audio-ai-2026/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Image, video and audio</category>
    <description>2026 guide to voice AI: TTS and transcription models, GPT-Live-1 vs Gemini 3.8 Live, latency budgets, AI music rights, voice cloning and compliance.</description>
  </item>
  <item>
    <title>AI cost per token in 2026: real cost, caching and routing</title>
    <link>https://slash-digital.io/en/insights/ai-costs/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/ai-costs/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Esteban Zuleta</dc:creator>
    <category>Strategy and regulation</category>
    <description>How AI is billed in 2026, cost per task versus price per token, savings levers with sourced figures, hidden costs and when local beats the API.</description>
  </item>
  <item>
    <title>LLM quantization explained: GGUF vs AWQ vs FP8 and FP4</title>
    <link>https://slash-digital.io/en/insights/quantization-guide/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/quantization-guide/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Local AI and open weights</category>
    <description>LLM quantization with data: GGUF, AWQ, EXL3, FP8, NVFP4 and MXFP4, how much quality you lose, the KV cache and how to validate a quant.</description>
  </item>
  <item>
    <title>RAG in 2026: long context, hybrid and agentic search</title>
    <link>https://slash-digital.io/en/insights/rag-2026/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/rag-2026/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Agents and engineering</category>
    <description>Does a 1M-token context make RAG obsolete? The 2026 pipeline: parsing, contextual retrieval, hybrid search, rerankers, evals, permissions and costs.</description>
  </item>
  <item>
    <title>AI guardrails in production: classifiers and policies</title>
    <link>https://slash-digital.io/en/insights/guardrails-production/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/guardrails-production/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Cybersecurity</category>
    <description>How to build AI guardrails that work: the layers, guard models compared (Llama Guard 4, Qwen3Guard, gpt-oss-safeguard), metrics and reference designs.</description>
  </item>
  <item>
    <title>AI-generated content: rights, licenses and provenance</title>
    <link>https://slash-digital.io/en/insights/ai-content-rights-provenance/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/ai-content-rights-provenance/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Image, video and audio</category>
    <description>AI-generated content in 2026: authorship, lawsuits, vendor indemnities, likeness, AI Act Article 50 labeling, C2PA, SynthID and a policy for brand teams.</description>
  </item>
  <item>
    <title>Self-hosted LLM server: Ollama, llama.cpp, vLLM, SGLang</title>
    <link>https://slash-digital.io/en/insights/local-llm-server/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/local-llm-server/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Local AI and open weights</category>
    <description>How to run LLMs in-house: Ollama, llama.cpp, vLLM and SGLang compared, batching and caching, security, sizing, and when an API is cheaper.</description>
  </item>
  <item>
    <title>LLM evals: how to test an LLM system before production</title>
    <link>https://slash-digital.io/en/insights/llm-evals/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/llm-evals/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Agents and engineering</category>
    <description>A practical guide to LLM evals: golden sets, LLM judges and their biases, human calibration, CI regression gates, tooling and the cost of running evals.</description>
  </item>
  <item>
    <title>AI supply chain security: models, packages, MCP</title>
    <link>https://slash-digital.io/en/insights/ai-supply-chain-security/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/ai-supply-chain-security/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Cybersecurity</category>
    <description>AI supply chain security: malicious model files, hallucinated packages, trojanized MCP servers, the controls that work and the EU CRA reporting duties.</description>
  </item>
  <item>
    <title>Generative AI in live brand activations and DOOH</title>
    <link>https://slash-digital.io/en/insights/generative-ai-brand-experiences/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/generative-ai-brand-experiences/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Image, video and audio</category>
    <description>How to use generative AI in brand activations and DOOH: the live pipeline, image models, biometric consent, AI Act labelling and success metrics.</description>
  </item>
  <item>
    <title>Small models and on-device AI: when they are enough</title>
    <link>https://slash-digital.io/en/insights/small-models-on-device/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/small-models-on-device/</guid>
    <pubDate>Sat, 26 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Rémy Sanchez</dc:creator>
    <category>Local AI and open weights</category>
    <description>Small language models and on-device AI in 2026: Apple Foundation Models, Gemini Nano, Gemma 4. What they handle, their limits and when to use the cloud.</description>
  </item>
  <item>
    <title>How Slash AI Lab chooses an AI model | Slash Digital Lab</title>
    <link>https://slash-digital.io/en/insights/how-we-choose-models/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/how-we-choose-models/</guid>
    <pubDate>Wed, 23 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Models</category>
    <description>Real production criteria: business Spanish, French, code, agents, cost, latency and privacy. The evaluation we run before every deployment.</description>
  </item>
  <item>
    <title>Fine-tuning vs RAG: the Slash AI Lab criterion</title>
    <link>https://slash-digital.io/en/insights/fine-tuning-vs-rag/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/fine-tuning-vs-rag/</guid>
    <pubDate>Wed, 23 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Agents and engineering</category>
    <description>When RAG, when fine-tuning and when a hybrid. No magic recipe: data, language and compliance constraints. The production criterion Slash applies.</description>
  </item>
  <item>
    <title>Pentest in Colombia: what to really test | Slash Digital Lab</title>
    <link>https://slash-digital.io/en/insights/pentest-colombia/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/pentest-colombia/</guid>
    <pubDate>Wed, 23 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Esteban Zuleta</dc:creator>
    <category>Cybersecurity</category>
    <description>Not a scanner PDF. Web apps and APIs with manual validation, an executive report and a retest. Launch offer in COP: $0 if there are no high findings.</description>
  </item>
  <item>
    <title>Do AI engines cite you? How to measure it</title>
    <link>https://slash-digital.io/en/insights/llm-visibility/</link>
    <guid isPermaLink="true">https://slash-digital.io/en/insights/llm-visibility/</guid>
    <pubDate>Wed, 23 Sep 2026 17:00:00 GMT</pubDate>
    <dc:creator>Jean Martial</dc:creator>
    <category>Strategy and regulation</category>
    <description>Measure whether ChatGPT, Claude or Perplexity cite your company correctly: 30 fixed prompts, 5 engines, 3 languages. Run on Slash first, in the open.</description>
  </item>
</channel>
</rss>
