AI in Marketing

Alibaba has launched its latest AI model, Qwen3.5, as part of its efforts to compete in China's rapidly evolving AI sector. The model is available in both open-weight and hosted API versions, offering improvements in performance and cost over previous iterations. Qwen3.5 is designed with native multimodal capabilities, enabling it to process text, images, and video simultaneously. It also supports new coding and agentic capabilities, making it compatible with open-source AI agents like those from OpenClaw. These AI agents can independently perform multi-step tasks with minimal supervision, a feature that has gained significant attention following the release of similar tools by American AI company Anthropic. Alibaba's competitors, including ByteDance and Zhipu AI, have also introduced upgraded models to enhance agent capabilities. The Qwen3.5 model boasts 397 billion parameters, fewer than its predecessor, but Alibaba claims it matches the performance of leading models from OpenAI, Anthropic, and Google DeepMind based on self-reported benchmarks. The model supports 201 languages and dialects, reflecting Alibaba's global ambitions. As Chinese AI companies prepare for potential disruptions to traditional internet business models, Alibaba plans to release more open-weight models during the Chinese New Year. Meanwhile, American AI companies are accelerating their development of agentic capabilities, with OpenAI and Google DeepMind making strategic moves in this area.

Next.

Lunar New Year has become a significant marketing opportunity for brands, with companies like Lush Cosmetics, Stanley 1913, and luxury fashion houses such as Burberry and Dior launching dedicated product lines and campaigns. These efforts often incorporate traditional symbols like the color red and horse motifs, reflecting the Year of the Fire Horse. While some brands have long celebrated the holiday, there is a noticeable increase in Western brand participation, driven by a growing interest in Asian culture. Amanda Lee Sipenock Fisher from Lush emphasizes the importance of authenticity, suggesting that successful campaigns involve collaboration with individuals who have a genuine understanding of the culture. Lush, for instance, begins planning months in advance, engaging employee resource groups to ensure cultural relevance in their products. Stanley 1913 also highlights the global nature of its efforts, with its APAC team providing cultural insights for their Lunar New Year collection. Despite not being the biggest sales event, the holiday allows brands to engage with communities celebrating it. Both Fisher and Stanley's Yu-Nien Chang note the potential for continued growth in this space, advocating for authentic cultural exchange and collaboration to maintain meaningful connections with consumers.

Meanwhile.

An online forum for current and former Homeland Security Investigations (HSI) officers serves as a platform for discussing concerns about Immigration and Customs Enforcement (ICE) operations. Members express discomfort with mass deportation efforts and criticize the interaction between federal agents and civilians. The forum, which includes over 2,000 members, features debates on the use of force, such as the controversial shootings of protesters Renee Good and Alex Pretti. Members are divided on these incidents, with some questioning the necessity of lethal force. The forum also highlights dissatisfaction with working conditions, including long hours and inadequate equipment. Agents criticize the allocation of resources, arguing that HSI's focus on administrative immigration enforcement detracts from more pressing criminal investigations. Complaints extend to the relationship between HSI and ICE's Enforcement and Removal Operations (ERO) teams, with some agents feeling that ERO is not effectively contributing to immigration work. Additionally, members discuss issues with ICE's technology and the agency's expansion across the U.S. Despite these challenges, ICE continues to grow its presence, indicating that detentions and deportations are likely to persist. The forum reflects a broader discontent among agents regarding the agency's direction and operational strategies.

On a different note.

The BBC reports that the AI Impact Summit in Delhi is focusing on guiding the AI revolution, with notable attendees including Bill Gates, despite initial uncertainty about his participation. The summit highlights the Global South's role in AI, contrasting with previous Western-dominated discussions. India, with significant tech hubs, is striving to advance its AI capabilities amid challenges like language barriers and limited resources compared to Western and Chinese counterparts. The summit aims to foster a more inclusive approach to AI governance, emphasizing people and progress. However, concerns persist about the event's effectiveness in addressing AI safety and responsibility, with some experts skeptical about meaningful outcomes. India's focus remains on leveraging AI for economic growth and digital sovereignty.

With that said.

Infosys has announced a partnership with Anthropic to develop AI agents for enterprise use, integrating Anthropic's Claude models into its Topaz AI platform. This collaboration aims to create systems capable of autonomously managing complex workflows in sectors like banking, telecoms, and manufacturing. The announcement was made at the AI Impact Summit in New Delhi, amid concerns that AI advancements could disrupt India's $280 billion IT services industry, which relies heavily on labor-intensive outsourcing. Earlier, Indian IT stocks fell after Anthropic launched AI tools for automating tasks in various roles. The partnership grants Infosys access to Anthropic's Claude models and developer tools, which will be used to enhance coding processes internally and for clients. Infosys reported that AI-related services contributed ₹25 billion, or 5.5% of its revenue in the December quarter. Anthropic's partnership with Infosys provides a pathway into regulated enterprise sectors, leveraging Infosys' expertise in industries like financial services. Anthropic also opened its first India office in Bengaluru, with India becoming its second-largest market. The timeline for deploying Claude-powered AI agents and financial terms of the deal remain undisclosed. This partnership mirrors similar collaborations, such as HCLTech's alliance with OpenAI to scale AI tools for enterprises.

After that.

San Francisco's AI startups are fostering a demanding work culture characterized by long hours and minimal downtime, reflecting broader pressures that may soon affect other industries. Workers in these startups often endure 12-hour days, seven days a week, driven by the intense pace of technological advancement and the fear of obsolescence. This environment is fueled by the rapid evolution of artificial intelligence, which has reshaped entry-level roles and increased job insecurity. High-profile tech leaders like Mark Zuckerberg and Elon Musk have openly discussed AI's potential to replace junior and mid-level positions, prompting companies to demand greater efficiency from their employees. The tech sector's shift from employee perks to heightened expectations has left many workers anxious about their job security and future roles. Despite these challenges, some employees remain motivated by the innovation in AI, dedicating extra hours to stay abreast of new developments. However, the pressure to constantly work and produce is seen as a warning sign for other sectors, as AI continues to transform the labor market. Economists predict significant changes in job structures, with AI potentially eliminating or altering a large percentage of roles in advanced economies, signaling a shift that could impact various industries globally.

Shifting gears.

The ARM Institute has announced a project call aimed at modernizing the organic industrial base (OIB) and its workforce, focusing on military supply chain technology. This initiative, exclusive to ARM members, seeks to address the outdated and poorly integrated equipment within OIB facilities that produce and sustain critical military technology and non-consumer goods. The project call, supported by the Office of the Secretary of Defense Manufacturing Technology (ManTech), emphasizes technical innovations and workforce solutions in areas such as digital operations technology, real-time manufacturing sensors for robotics, AI robotic process planning, in-situ quality checks, reduced operator exposure, cost reduction, and mobile and large surface automation. ARM members can access detailed project call information through the ARM member community, with whitepaper submissions due by March 2 and project performance deadlines set for October 1. The ARM Institute expects funding for selected projects to surpass previous levels, requiring submissions to propose concepts that result in deployable prototypes. Successful teams are likely to have existing relationships with OIB facilities, and submissions must include a letter of commitment from a military service or OIB site. The ARM Institute encourages joint technical and workforce proposals, with a focus on training the workforce on technology changes. A webinar on February 19 will provide further details on the project call and submission process.

In other news.

AI agents are reshaping B2B marketing by becoming the primary decision-makers in vendor selection, with projections indicating that 90% of B2B purchases will be managed by AI within three years. This shift is driven by AI's ability to synthesize vast amounts of data, forming preferences based on user reviews, technical documentation, and integration capabilities rather than traditional search methods. The rise of AI agents means that being the top recommendation is crucial for businesses, as these agents influence purchasing decisions significantly. Companies like Salesforce, ServiceNow, and Microsoft are developing autonomous systems to handle procurement and vendor selection, emphasizing the importance of being favored by these AI agents. The emergence of agent-mediated commerce is evident with platforms like ChatGPT, which now allows direct purchases within conversations, and Shopify's "Agentic Storefronts," which manage product visibility across multiple AI platforms. This trend is not limited to consumer markets; it is rapidly transforming the B2B sector, where AI recommendations can reach millions of potential buyers. As AI agents continue to evolve, businesses must focus on product quality, data structure, and platform integrations to ensure they are recommended by these systems. The timeline for this transformation is imminent, with significant changes expected by 2025-2026, making it essential for companies to adapt quickly to maintain competitiveness.

Elsewhere.

Sumit Virmani, CMO of Infosys, emphasizes the evolving role of CMOs within the C-suite, highlighting the strategic importance of AI in expanding their influence beyond traditional marketing roles. Speaking at the 2026 World Economic Forum in Davos, Virmani argues that CMOs must engage in cross-functional business discussions, using AI-driven insights to contribute to enterprise-wide decisions. He stresses the need for CMOs to redefine marketing success as driving transformation and growth, rather than focusing solely on creative campaigns. This shift requires CMOs to develop skills in data analytics and business operations. Virmani also advises CMOs to maximize the impact of attending high-level forums by focusing on brand visibility, customer relationships, and competitive intelligence gathering.

Finally.

By late 2025, the enterprise AI sector had transitioned from standard retrieval-augmented generation (RAG) systems to autonomous agents due to the high failure rate of RAG systems. These agents, while solving reliability issues, introduced new security challenges, including the autonomous execution of malicious instructions. The shift to agentic AI was driven by the need to address the limitations of RAG systems, which struggled with large datasets and often produced inaccurate results due to infrastructure constraints. This led to the development of self-reflective RAG, which evaluates the usefulness of data before processing it, and corrective RAG, which uses an evaluator model to ensure data accuracy. However, the rise of agentic systems has increased security risks, such as indirect prompt injection, memory poisoning, and agentic denial of service (DoS). These vulnerabilities allow attackers to exploit AI systems without direct interaction, leading to unauthorized actions and data breaches. Real-world exploits, like the EchoLeak vulnerability in Microsoft Copilot, have demonstrated the potential for significant data exfiltration. To counter these threats, companies like NVIDIA and Lakera AI are developing comprehensive defense strategies, including the unified safety framework, which emphasizes securing the entire system rather than just the language model. This approach involves using guardian agents to monitor and correct the primary agent's actions in real time.