Mergers and acquisitions continue to spread in the global security market. The industry is also making moves to respond to AI security threats. Browser security has emerged as a strategic focal point.

Cloud security firm Zscaler has acquired browser security specialist SquareX. SquareX has technology that blocks browser-based threats such as phishing and session hijacking. It detects threats in real time by monitoring in-browser behavior without relying on URL reputation or network inspection. It provides security through a lightweight extension that runs on existing browsers such as Chrome and Edge, without requiring a separate enterprise browser.

Zscaler acquires browser security firm SquareX, built to run on existing browsers.

Security firm Sophos acquired UK security startup Arco Cyber as part of an expansion of AI-based governance. Sophos plans to use the deal to help organisations without dedicated security leaders build risk-response capabilities at the chief information security officer level. AI agent security startup Keycard is acquiring authentication management specialist Anchor.dev. Keycard CEO Ian Livingstone said, "As AI agents gain more and more authority, security and trust issues are becoming important," and added, "This acquisition will further strengthen our AI agent security technology."

Sophos buys UK security firm Arco, saying it will use AI to fill the CISO gap. Keycard acquires Anchor, expanding beyond AI agent ID security toward an integrated management platform.

In AI, industry moves to expand beyond large language models (LLMs) toward world models that understand the real world are accelerating.

Alibaba unveiled an AI model called RynnBrain to help robots understand and identify their surroundings. RynnBrain targets physical AI, including robotic technologies such as self-driving cars. AI video generation startup Runway raised $315 million in a Series E funding round. South Korean AI model startup TrillionLabs also developed a mobile world model, gWorld-32B, that learns causal relationships in the surrounding environment to simulate future changes.

Alibaba launches robot-focused AI model RynnBrain, expanding into physical AI. Runway raises $315 million, accelerating world model development. TrillionLabs develops mobile world model gWorld-32B.

Moves by tech companies in South Korea and abroad to develop or use AI continued this week.

OpenAI will introduce advertising for ChatGPT users on its free and low-cost plan, Go. The company stressed that "ads do not affect ChatGPT responses, and users' conversation content is not shared with advertisers" to minimise changes to user experience from the introduction of ads. With OpenAI unveiling its latest AI model, GPT-5.3 Codex, attention is on whether it can to some extent keep Anthropic in check after Anthropic jumped ahead in the coding AI race.

ChatGPT introduces ads, applying them to free and low-cost plans. OpenAI steps up its coding push, as focus turns to whether it can keep Anthropic in check.

Goldman Sachs is developing AI agents in partnership with AI startup Anthropic. Goldman Sachs is building automation systems with Anthropic AI engineers and is pursuing a strategy of actively using AI, particularly in accounting and customer screening.

Goldman Sachs pursues accounting and compliance automation based on Anthropic AI models.

ByteDance has released an AI-based video generation model, Seedance 2.0. Seedance 2.0 is getting responses among users calling it a "game changer", the report said, citing features such as multi-scene processing, audio and dialogue synchronisation and multilingual support.

ByteDance unveils AI video generation model Seedance 2.0.

Kantar Korea, the South Korean unit of UK-headquartered global marketing data and analytics specialist Kantar, is launching a GEO service that analyses how brands are perceived and chosen by consumers in a generative AI environment and presents growth strategies.

Kantar Korea starts GEO service for brand strategy in the generative AI era.

Cadence Design Systems announced ChipStack, an AI "super agent" that can be used in semiconductor design and verification.

Cadence announces ChipStack, an AI agent aimed at semiconductor design.

Former GitHub CEO Thomas Dohmke (토마스 돔케) founded development tools startup Entire and raised $60 million in seed funding at a $300 million valuation. Entire is developing an open-source tool to effectively manage code generated by AI agents. IBM invested in design-to-code startup Anima to transform how companies design digital products for the AI era. Community platform Reddit is actively pursuing acquisitions of AI and adtech companies. Reddit CFO Andrew Vollero (앤드류 볼레로) said, "We will expand our user base by acquiring companies that can leverage Reddit's scale." Reddit is speeding up efforts to make AI-based search a new revenue model and combine it with traditional search.

Former GitHub CEO raises $60 million, entering AI development tools market. IBM invests in generative AI app design startup Anima. Reddit moves to strengthen AI and ad technology, pushing additional acquisitions. Reddit positions AI search as the next revenue model, seeking to combine it with traditional search.

A new AI platform, AI.com, founded by Crypto.com's co-founder and CEO Kris Marszalek (Kris Marszalek), is launching a consumer "autonomous AI agent" service. The AI.com service supports anyone creating their own personal AI agent with just a few clicks. WordPress launched a new connector that links to Anthropic AI chatbot Claude. AI lifelog service company HeyDiary formally launched its voice-based diary app, HeyDiary. Visual communication platform Canva unveiled Canva AI Connector, which directly links brand assets to AI assistant tools such as ChatGPT. Global AI company DeepL launched the DeepL Voice API, which provides real-time speech recognition and translation.

Crypto.com's Kris Marszalek launches a personal AI agent creation service. WordPress links with Anthropic's Claude, supporting use of site data. HeyDiary formally launches voice-based AI diary app. Canva unveils AI Connector, enabling brand-tailored design creation with ChatGPT. DeepL launches real-time speech translation API.

On forecasts that AI will threaten the existing software market, Databricks CEO Ali Ghodsi said AI is not replacing SaaS, but is instead increasing usage. In Databricks' case, database use is surging through its generative AI interface, Genie, he said. Databricks raised $5 billion at a $134 billion valuation.

Databricks CEO says "SaaS is not dead...the key is shifting to natural-language interfaces". Databricks raises $5 billion at a $134 billion valuation.

In the LLM-based AI market, Anthropic has been called part of the big three with OpenAI and Google, but it had a strong No. 3 image. But that has changed. Anthropic has sharply raised its standing in the tech industry since the start of the year. The shift gained momentum after Anthropic unveiled its AI tool Claude Codework in January. Claude Codework, an AI tool that lets even non-developers build work tools, is still a preview version, but it has helped elevate Anthropic into the ranks of AI issue-makers that surpass OpenAI and Google, aligning with a narrative that it could threaten the existing software market.

[Tech Inside] The revolt of Anthropic, the "AI No. 3", shook global stock markets.

SaaS was not the only area affected by Anthropic. There is also analysis that Anthropic's new product launches are putting direct pressure on the vertical AI ecosystem. The view is that foundation model (FM) developers such as Anthropic are strengthening expertise across various domains to expand platforms and increase revenue, bringing companies that specialise in vertical AI into their direct sphere of influence.

Foundation model verticalisation accelerates, raising questions over whether vertical AI can survive.

Moves by AI developers, major retailers and payment service companies targeting AI shopping are accelerating. Even competitors appear to be joining forces when needed. Unusual alliances among companies taking on the world's largest e-commerce company, Amazon, are spreading. The report says companies are actively pursuing cooperation even with firms that compete with them in certain businesses, because taking on Amazon comes first.

Anti-Amazon "AI commerce alliance" spreads, with rivals joining hands when needed.