News You Can Use

DeepSeek R1 is a newly released open-source LLM designed to deliver strong reasoning capabilities and cost-efficient performance, positioning itself as a competitor to the big players. Unlike others, DeepSeek claims to have superior mathematical reasoning and improved instruction-following abilities, making it particularly useful for professional and enterprise applications. Benchmarks indicate it performs competitively in reasoning and problem-solving tasks while being open source.

The emergence of DeepSeek presents a potential alternative to proprietary AI models, allowing AG to explore cost-effective, customisable AI solutions while reducing dependency on closed-source systems. Given its claimed strengths in reasoning, it could be particularly useful in contract analysis, risk assessment and contract family reviews. Testing its effectiveness in real-world legal scenarios could inform whether open-source models can match or exceed proprietary solutions for AG's use cases. The challenge will be getting a safe environment set up to execute on this.

This report provides a deep dive into the ALSP market's rapid expansion. The ALSP sector now stands at $28.5 billion, growing at an 18% compound annual growth rate (CAGR). Notably, over 50% of corporate law departments now rely on ALSPs for services like eDiscovery, litigation support, and legal operations consulting. However, the market is splitting - while some law firms embrace ALSPs as partners, others remain resistant. A key disruptor is GenAI, which 35% of law firms and 40% of corporate legal departments see as making ALSPs more attractive. GenAI allows ALSPs to offer faster, cost-efficient services, potentially eroding law firms' competitive advantage unless they adopt similar technologies. The long-term impact remains uncertain, as firms may build in-house AI expertise to offset reliance on ALSPs.

The findings highlight the growing necessity for legal innovation. As ALSPs integrate GenAI to optimise workflows, traditional firms that fail to evolve risk losing ground. For AG, this underscores the urgency of leveraging AI-driven efficiencies in legal services while maintaining high-quality, strategic advisory roles that ALSPs cannot easily replicate. It also backs up our recent move into the ALSP market with our MLS offering. Monitoring how clients adopt ALSPs could inform AG's strategy in offering AI-augmented solutions that align with shifting client expectations.

The next generation of AI assistants is emerging, designed to handle specific tasks and workflows rather than just providing conversational responses. These assistants integrate with external tools, APIs, and knowledge sources to deliver more actionable outputs. OpenAI's Operator is a customisable AI-powered agent for businesses, aimed at improving information retrieval with better-sourced responses, reducing research time. While Perplexity's Assistant focuses on search and knowledge retrieval, synthesising information from multiple sources to provide real-time, cited answers.

These advancements highlight the broader trend of AI shifting from general-purpose models to more specialised, function-driven assistants. For contract analysis and due diligence, AI-driven workflow automation could help extract key clauses, identify risks, and streamline review processes - potentially integrating with tools we are currently building. As AI assistants become more autonomous, there is a potential to explore how APIs and external integrations could allow us to embed AI into existing legal workflows, either through 3rd parties or through building on solutions such as Power Automate or Autologyx.