Meta has agreed to acquire Singapore-based AI startup Manus for approximately $2 billion in January 2026. The company, founded by Chinese entrepreneurs under parent company Butterfly Effect, is known for autonomous AI agents that handle coding and market research, and will continue operating independently while integrating technology across Meta's platforms.

Source: AI Insider

Microsoft's unified Agent Framework, merging AutoGen's multi-agent patterns with Semantic Kernel's enterprise features, is scheduled for general availability in Q1 2026. The framework offers native Azure integration, multi-language support (C#, Python, Java), production SLAs, and formal support contracts for enterprises in the Microsoft ecosystem.

Source: Medium

Microsoft is launching agentic features in Dynamics 365 with a preview scheduled for February 2026, enabling retailers to build autonomous customer experiences. The announcement comes amid industry pressure as 57% of leaders fear AI content oversupply while 62% of organizations remain in experimental phases with agentic AI.

Source: AI Agent Store

A critical remote code execution (RCE) vulnerability was discovered in the open-source AI agent framework Langflow, which was also exploited in the wild in 2025. Additional RCE vulnerabilities were found in major AI inference server frameworks from Meta, Nvidia, Microsoft, and open-source projects like vLLM and SGLang.

Source: CSO Online

Anthropic's Model Context Protocol (MCP) has become the standard for how LLMs interact with external data sources and applications. OpenAI and Microsoft have publicly embraced MCP, and Anthropic recently donated it to the Linux Foundation's new Agentic AI Foundation. Google has also begun standing up its own managed MCP servers.

Source: TechCrunch

OpenAI announced it will retire several older AI models including GPT-5 (Instant and Thinking), GPT-4o, GPT-4.1, GPT-4.1 mini, and o4-mini from ChatGPT on February 13, 2026. The company cites low usage and shift toward newer models, with only 0.1% of users still choosing GPT-4o daily as vast majority has shifted to GPT-5.2.

Source: Business Standard

OpenAI confirmed retirement of GPT-4o on February 13, 2026, alongside previously announced retirement of GPT-5 (Instant and Thinking). The company states GPT-5.2 has now lived up to expectations and incorporated feedback from GPT-4o users who preferred its conversational style and warmth.

Source: BleepingComputer

OpenAI officially introduced GPT-5 as their best AI system yet, featuring significant leaps in intelligence across coding, math, writing, health, and visual perception. The model shows particular improvements in complex front-end generation, debugging larger repositories, and can create beautiful websites and apps with single prompts.

Source: OpenAI

OpenAI released GPT-5.2 as the most capable model series for professional knowledge work, achieving state-of-the-art performance on GDPval benchmark where it outperforms industry professionals on 70.9% of knowledge work tasks across 44 occupations. The model is rolling out today starting with paid plans.

Source: OpenAI

OpenAI released GPT-5.2-Codex as the most advanced agentic coding model for complex software engineering, featuring improvements in long-horizon work, large code changes, Windows environments, and significantly stronger cybersecurity capabilities. It's available today for paid ChatGPT users with API access coming in weeks.

Source: OpenAI

CISA added CVE-2026-1281 to its Known Exploited Vulnerabilities catalog with an unusually short February 1 remediation deadline for federal agencies. CISA added five new vulnerabilities to its Known Exploited Vulnerabilities catalog on January 26 based on evidence of active exploitation: CVE-2018-14634 (Linux Kernel Integer Overflow), CVE-2025-52691 (SmarterTools SmarterMail Unrestricted Upload), CVE-2026-21509 (Microsoft Office Security Feature Bypass), CVE-2026-23760 (SmarterTools SmarterMail Authentication Bypass), and CVE-2026-24061 (GNU InetUtils Argument Injection).

Source: Innovate Cybersecurity

Published 2026-02-02, The MediaTek Product Security Bulletin contains details of security vulnerabilities affecting certain MediaTek chipsets. Device OEMs have been notified of all the issues and the corresponding security patches for at least two months before publication. At this time, we are not aware of any active exploitation of these vulnerabilities in the wild.

Source: MediaTek

Ivanti has released provisional patches that fix two critical code injection vulnerabilities in Endpoint Manager Mobile (EPMM), one of which (CVE-2026-1281) has been exploited in zero-day attacks and has been added to CISA's Known Exploited Vulnerabilities catalog. Both CVE-2026-1281 and CVE-2026-1340 are code injection flaws affecting EPMM's In-House Application Distribution and Android File Transfer Configuration features. They may allow unauthenticated attackers to achieve remote code execution (RCE) on vulnerable on-premises EPMM installations.

Source: Help Net Security

The US Cybersecurity and Infrastructure Security Agency added CVE-2026-21509 to its Known Exploited Vulnerabilities catalog and ordered US federal civilian agency to address the flaw by February 16, 2026. Microsoft released emergency Office security updates to fix a zero-day vulnerability (CVE-2026-21509) actively exploited by attackers. CVE-2026-21509 stems from reliance on untrusted inputs in a security decision in Microsoft Office, which allows unauthorized attackers to bypass a security feature (OLE mitigations in Microsoft 365 and Microsoft Office) locally.

Source: Help Net Security

Cybersecurity and Infrastructure Security Agency (CISA) has added CVE-2026-20805 to its Known Exploited Vulnerabilities (KEV) catalog. Federal Civilian Executive Branch (FCEB) agencies are now required to apply the corresponding updates no later than February 3, 2026, highlighting the urgency of remediation. The actively exploited flaw, tracked as CVE-2026-20805, has been classified by Microsoft as an important-severity information disclosure vulnerability affecting the Windows Desktop Window Manager (DWM). The issue allows a locally authorized attacker to access sensitive information by abusing weaknesses in how DWM handles system data, potentially exposing details that should otherwise remain protected.

Source: SOC Prime

Microsoft recommends using GitHub Copilot, which provides advanced suggestions and inline completions, while GitHub Copilot Free is limited to 2,000 code completions and 50 chat or 'agent mode' requests per month. For heavy users, Microsoft is now effectively moving toward a model that requires $120 to $228 per year.

Source: Visual Studio Magazine

This will probably be the last major VS Code update until February 2026. Microsoft skips the December release cycle. Ten new agent features shipped in this update. VS Code includes a setting called YOLO (you only live once) that disables manual approval for all tools in all workspaces. The setting description warns that 'this feature disables critical security protections and makes it much easier for an attacker to compromise the machine'.

Source: DevOps.com

Release date: January 08, 2026. VS Code now supports Agent Skills, allowing you to teach the coding agent new capabilities and provide domain-specific knowledge. Agent Skills are folders of instructions, scripts, and resources that GitHub Copilot can load when relevant to perform specialized tasks. This year, we managed to reduce our open issues by nearly 6,000 and triaged over a thousand more.

Source: VS Code Official Release Notes

This plan captures our work in January 2026. This is a 4-week iteration. We will ship in early February. The plan includes various bug fixes and improvements including Python environment extensions and chat code review commands.

Source: Microsoft VS Code GitHub

AI-based VS Code forks recommended unclaimed extensions, allowing malicious uploads in Open VSX and risking developer systems. This represents a security vulnerability in the VS Code ecosystem where threat actors can exploit missing extensions.

Source: The Hacker News

AWS announced the general availability of the AWS European Sovereign Cloud on January 14, 2026, a new independent cloud entirely located within the EU. Amazon plans to invest more than €7.8 billion in this sovereign cloud infrastructure in Germany and will expand to Belgium, Netherlands, and Portugal with new AWS Local Zones.

Source: Business Wire

AWS launched Amazon EC2 G7e instances with NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, delivering up to 2.3x better inference performance compared to G6e instances. Amazon Corretto received quarterly security updates for versions 25.0.2, 21.0.10, 17.0.18, 11.0.30, and 8u482.

Source: AWS News Blog

AWS expanded availability of Amazon EC2 R8i and R8i-flex instances to additional regions in January 2026. These instances are up to 30% faster for PostgreSQL databases, up to 60% faster for NGINX web applications, and up to 40% faster for AI deep learning recommendation models compared to R7i instances.

Source: AWS

AWS announced an investment of up to $50 billion to build AI and supercomputing infrastructure for U.S. government agencies starting in 2026. The investment will add nearly 1.3 gigawatts of AI and supercomputing capacity across AWS Top Secret, AWS Secret, and AWS GovCloud regions.

Source: Fox Business

AWS announced new partner programs for 2026 including up to $50k Marketing Development Funds for Amazon Connect implementations launching January 2026, expansion of the Partner Greenfield Program, and new MSP benefits. AWS Partners can generate $7.13 in revenue for every $1 of AWS services sold.

Source: AWS Partner Network Blog

Two additional critical vulnerabilities were disclosed in Next.js: CVE-2025-55184 (CVSS 7.5) allowing denial of service attacks and CVE-2025-55183 (CVSS 5.3) enabling source code exposure. These follow the initial React2Shell vulnerability and require immediate patching.

Source: Next.js Official Blog

React disclosed React2Shell (CVE-2025-55182) with CVSS 10.0 severity, allowing remote code execution in React Server Components. All Next.js 15.x and 16.x users must upgrade immediately, along with patches for additional DoS and source code exposure vulnerabilities.

Source: React Official Blog

Following last week's critical vulnerability, two additional CVEs were disclosed on December 11, 2025. CVE-2025-55184 causes server hang attacks, while CVE-2025-55183 leaks Server Function source code. All App Router sites are vulnerable and must upgrade again.

Source: Netlify Changelog

Security researchers warned about critical vulnerabilities in React Server Components and Next.js with CVSS 10.0 severity scores. The flaws enable unauthenticated remote code execution and have nearly 100% success rate in exploitation attempts.

Source: Cybersecurity Dive

New vulnerabilities CVE-2025-55184 and CVE-2025-67779 allow denial of service attacks, while CVE-2025-55183 enables source code disclosure. Previous React2Shell patches are insufficient, requiring additional upgrades even for previously patched environments.

Source: Orca Security