Akamai, MuleSoft Unify API Security

Key Takeaways

• Akamai and MuleSoft expanded their partnership on September 10, 2026, to unify security for AI agents and APIs.
• 87% of organizations faced API security incidents in 2025, driving the need for better security tools.
• More than 20 companies are already using the Akamai and MuleSoft integration to find hidden security gaps.
• AI hardware sales in Shenzhen's Huaqiangbei district grew by more than 55% from January to July 2026.
• DeepSeek AI released DeepSeek-V4.1-Flash with a 1 million token context window and 45 trillion multimodal tokens training data.
• Harvard Business Impact and CodeSignal partnered to provide hands-on AI courses with an AI tutor named Cosmo.
• Patrick Gruhn, CEO of Perpetuals, will lead a three-year doctoral project with EIM to develop AI-driven trading models.
• The White House finalized a voluntary AI testing framework in August but has not publicly released it.
• OpenAI has submitted models for review under the White House's voluntary AI testing framework.
• Naver is building an in-house security AI using 4,000 GPUs after access to Anthropic's Claude was disrupted.
• China warned Japan against plans to use AI-equipped tactical satellites for counterstrike capabilities.

Akamai and MuleSoft join forces to secure AI agents

Akamai and MuleSoft announced an expanded partnership on September 10, 2026, to unify security for AI agents and APIs. This collaboration addresses a major risk where 87% of organizations faced API security incidents in 2025. The new system connects MuleSoft Exchange with Akamai to share data about API specifications and threat risks. More than 20 companies are already using this integration to find hidden security gaps. Future updates planned for late 2026 will add better tools for managing AI environments.

Akamai and MuleSoft combine tools to fix API risks

Akamai and MuleSoft have merged their technology to improve security for AI agents and Model Context Protocol servers. The integration links Akamai API Security with MuleSoft Agent Fabric to create a two-way flow of information. This helps companies spot undocumented endpoints and monitor traffic that might be unsafe. Oz Golan from Akamai noted that teams cannot protect APIs they cannot see. Karl Albertsen from MuleSoft added that this partnership strengthens governance across digital business assets.

Akamai and MuleSoft team up to stop API threats

Akamai Technologies and MuleSoft expanded their collaboration to help businesses secure unmonitored APIs and shadow systems. This new partnership aims to reduce the risk of data breaches by improving API discovery and security analytics. The companies will work together to provide AI-powered measures against emerging cyber threats. Akamai is a leading provider of cloud security solutions that protect enterprise applications. This joint effort helps organizations stay ahead of growing security risks in the digital world.

Shenzhen Huaqiangbei sees huge AI hardware sales rise

The Huaqiangbei district in Shenzhen reported a rapid increase in AI hardware sales ahead of APEC 2026. Sales of AI products grew by more than 55% from January to July 2026 compared to the previous year. AI glasses sales doubled, while drone and robotics sales jumped by 60% to 70%. These AI items now make up 61% of all electronics sales in the district. The area handles over 1 billion parcels annually and serves customers in 183 countries and regions.

DeepSeek AI launches V4.1-Flash with massive context

DeepSeek AI released a new model called DeepSeek-V4.1-Flash that supports a context window of 1 million tokens. This update uses a technique called Cross-Layer Attention Reuse to save memory and speed up processing. The model uses FP4 KV cache compression to reduce the size of stored data significantly. It was trained on 45 trillion multimodal tokens and matches the knowledge of larger models while using fewer parameters. The new version scored higher than previous models on coding and automation benchmarks.

Harvard Business Impact partners with CodeSignal for AI skills

Harvard Business Impact and CodeSignal joined forces to bring hands-on AI and essential skills courses to universities worldwide. This partnership helps students practice using real AI tools and develop skills like prompt engineering and navigating change. Students receive immediate feedback from an AI tutor named Cosmo while working through realistic workplace scenarios. The initiative aims to close the gap between what students learn in school and what employers need in the AI era.

Perpetuals and EIM launch three-year doctoral AI trading project

Perpetuals and the European Institute of Management (EIM) are starting a multi-year project to develop AI-driven trading models. More than 20 doctoral researchers will work together using Perpetuals proprietary market data to study behavioral finance and model risk. Patrick Gruhn, CEO of Perpetuals, will help lead this group which includes many current employees of the company. The research is expected to run for at least three years to create real-world applications for trading.

White House AI testing framework remains secret amid safety concerns

The White House has not publicly released its voluntary framework for testing frontier AI models finalized in August. Companies like OpenAI have submitted models for review, but the government does not require firms to share the results of these evaluations. Critics argue that keeping the framework hidden creates a lack of transparency regarding safety standards. The administration believes the U.S. must maintain global dominance in artificial intelligence to compete with China.

Naver builds in-house security AI after Claude access disruption

Naver decided to create its own AI security model after export controls blocked access to Anthropic's Claude service. The company realized it could not rely on overseas models to understand domestic threats or meet local security needs. Naver joined a government consortium with 32 other organizations to develop a specialized cybersecurity AI foundation model. This project will use 4,000 of Naver's GPUs and aims to launch a demonstration model by December.

China warns Japan over AI satellite space militarization plans

China has warned that Japan's plan to use AI-equipped tactical satellites poses a serious threat to regional security. Japan intends to develop these satellites to speed up targeting decisions and strengthen its counterstrike capabilities. The defense ministry plans to build a prototype satellite with domestic companies and conduct ground tests in the coming years. Funding for this project is included in Japan's draft defense budget request for the 2027 financial year.

Sources

• Akamai and MuleSoft Unify Runtime Enforcement Across Agent Fabric – Bridging the Gap Between Security Policy and AI Orchestration
• Akamai And MuleSoft Expand Integration To Unify API Security And AI Agent Governance
• Akamai Expands MuleSoft Collaboration to Address API, AI Security Risks
• Shenzhen Huaqiangbei Reports Rapid AI Hardware Growth Ahead of APEC 2026
• DeepSeek AI Released DeepSeek-V4.1-Flash with 1M Context, FP4 KV Cache, and Cross-Layer Attention Reuse
• CodeSignal and Harvard Business Impact Bring Hands-On AI and Essential Skills Development to Universities Worldwide
• More than 20 doctoral researchers join an AI trading project expected to run at least 3 years
• White House framework for testing AI models remains hidden as concerns about threats mount
• Naver Turns to In-House Security AI After Claude Access Disruption
• Japan’s AI satellites and other space ambitions ‘a serious threat’, China warns