Intellect-Partners

Tag: machinelearning

Categories
Others

AI Patent Platform Fearn Secures $5.5M Seed Round to Automate Drafting

AI Patent Platform Fearn Secures $5.5M Seed Round to Automate Drafting

San Francisco-based intellectual property startup Fearn has announced the completion of a $5.5 million seed funding round to expand its AI-native patent drafting platform.

The round was led by Kindred Ventures, with participation from Andreessen Horowitz’s a16z speedrun startup accelerator, Designer Fund, and Essence VC. Prior to this institutional round, the company operated via founder self-funding.

The Founders and the Logic of Automation

Fearn was founded in 2025 by CEO Han Kim and CTO Angela Gao, who met as graduate students at Caltech. The platform’s architectural focus stems directly from the co-founders’ specialized backgrounds:

  • Han Kim: Previously prosecuted patent applications across software, life sciences, and mechanical arts as a scientific analyst at Morrison & Foerster, while researching bio-inspired neural algorithms during his Ph.D. track at Caltech.
  • Angela Gao: Completed a Ph.D. in computing and mathematical sciences at Caltech, specializing in physics-aligned generative models, alongside previous model development work at Google Research.

Kim noted that his experience in Big Law highlighted systemic inefficiencies in the traditional patent pipeline, which is frequently slow, cost-prohibitive, and anxiety-inducing for engineers worried about technical details being misinterpreted.

“I noticed a lot of the tasks I was doing could be automatable, but obviously I couldn’t automate them. You’re not really allowed to in those sorts of settings and environments,” Kim stated, highlighting the strict procedural friction within traditional law firms that inspired him to build an external automation solution.

How the Multi-Model Stack Works

Unlike general-purpose generative AI tools or simple API wrappers, Fearn is built from the ground up as a fully data-sovereign, AI-native platform. The coordinates a specialized multi-model stack:

  • Bespoke Model Ensemble: The platform utilizes dozens of hypercompact, specialized models, combining proprietary code, fine-tuned open-source models, and symbolic non-LLM systems built from scratch.
  • Data Sovereignty: Fearn hosts 100% of its own model stack internally. It makes zero application programming interface (API) calls to third-party model developers, completely removing the public-disclosure and data-egress risks that typically restrict enterprise IP teams from leveraging generative AI.
  • Hallucination Resistance: By training its custom architectures on highly curated, hand-corrected, and hand-labeled intellectual property datasets, Fearn creates audit trails engineered to guarantee compliance with patent office requirements and eliminate the factual errors common in large language models.

Once the application and automated labeled figures are ready, corporate research teams or solo inventors can choose to file the paperwork independently or hand it off to external counsel for final strategic review. Fearn charges a flat, predictable fee of $2,000 per patent draft, cutting traditional preparation timelines down significantly.

Future Plans and the Legal Tech Boom

With a lean team of fewer than 10 people, Fearn plans to deploy the capital injection primarily toward technical hiring, infrastructure expansion, and offsetting computational overhead.

Looking forward, the company intends to scale its features to assist inventors throughout the entire end-to-end patent prosecution lifestyle. This includes expanding automated systems to handle office action responses and any procedural workflow tied directly to a U.S. Patent and Trademark Office (USPTO) registration number.

Fearn’s successful seed round emphasizes an accelerating streak of legal tech investments by Andreessen Horowitz. The firm’s recent IP and legal portfolio expansion includes:

  • Leading patent automation startup Stilta’s seed round.
  • Anchoring multiple massive funding rounds for legal AI platform Harvey.
  • Backing litigation-focused developer Eve across two distinct rounds.
  • Leading the pre-seed round for communication security provider ZeroDrift.
Categories
Computer Science Electronics

Microsoft’s Explainability Patent Paves the Way for Trustworthy AI

In the rapidly evolving landscape of Artificial Intelligence, the pursuit of groundbreaking innovation often intersects with the critical need for transparency and trust. A recent patent application from tech giant Microsoft, focusing on a “generative AI for explainable AI,” underscores this crucial intersection, highlighting a significant step towards demystifying how AI models arrive at their conclusions. For businesses navigating the complexities of AI adoption, understanding the implications of such intellectual property is paramount.

Two Minds Are Better Than One: A Novel Approach to AI Explanations

Microsoft’s innovative approach posits that the best way to understand one generative AI model is to employ another. This patent application reveals a system designed to illuminate the inner workings of machine learning outputs, providing users with much-needed clarity on the ‘why’ behind an AI’s decision.

Imagine an AI system being queried: “Why was this loan approved (or denied)?” Microsoft’s proposed technology doesn’t just offer a single answer. Instead, it meticulously analyzes the input data (the loan application), alongside relevant historical data, user preferences, past explanations, and even subject matter expertise. This comprehensive analysis generates multiple potential explanations for the AI’s output.

But the innovation doesn’t stop there. Crucially, the system then leverages a second generative AI model to rank these potential explanations based on their relevance and clarity. This multi-layered approach aims to deliver not just an explanation, but the most pertinent explanation, fostering genuine understanding and confidence in AI-driven outcomes.

The Imperative of Explainable AI (XAI) in Enterprise Adoption

As Microsoft succinctly states in its filing, Explainable AI (XAI) “helps the system to be more transparent and interpretable to the user, and also helps troubleshooting of the AI system to be performed.” This statement resonates deeply with the challenges faced by enterprises deploying AI today.

The race to build and deploy advanced AI is undeniable, yet persistent issues like algorithmic bias and “hallucinations” (AI generating false information) continue to erode trust and pose significant liability risks. Without robust monitoring and a clear understanding of AI decision-making processes, the promise of AI can quickly turn into a peril.

This is precisely why responsible AI frameworks are gaining traction across industries. A recent McKinsey report highlighted this trend, revealing that a majority of surveyed companies are committing substantial investments – over $1 million – into responsible AI initiatives. The benefits are clear: enhanced consumer trust, fortified brand reputation, and a measurable reduction in costly AI-related incidents.

Protecting Your AI Innovations: The Role of Intellectual Property

For a patent intellectual property firm, Microsoft’s move is a powerful signal. As companies like Microsoft push the boundaries of AI, protecting the underlying methodologies and novel applications becomes critical. Patents like this one not only secure a competitive advantage in the burgeoning AI market but also provide a shield against potential liabilities that arise from AI’s complex and sometimes opaque nature.

By actively researching and patenting explainable and responsible AI technologies, Microsoft is not just aiming for a lead in the “AI race”; it’s strategically building a foundation of trust and accountability. This proactive approach to intellectual property in AI, particularly around explainability, could significantly bolster a company’s reputation and safeguard its innovations against future challenges.

For businesses developing or deploying AI, understanding the nuances of AI patents and the strategic importance of explainability is no longer optional – it’s a fundamental pillar of responsible and successful AI integration.

Categories
Computer Science

IP in the Age of AI: Who Owns the Algorithm?

In an era where artificial intelligence systems are designing new drugs, composing symphonies, and even writing code, the lines between creator and machine are becoming blurred. As AI continues to infiltrate nearly every industry, the question of intellectual property (IP) ownership is more relevant—and more complex—than ever before.

But when it comes to algorithms, especially those designed by or with the help of AI, who really owns the rights?

A Shifting Landscape

Traditionally, intellectual property laws were crafted with human inventors, artists, and developers in mind. The statutes assume a direct line between a person and their creation. But now that machines can “create” based on training data and optimization, the framework no longer fits as neatly.

Take, for example, a neural network trained to generate new software code. If a developer sets up the AI model, feeds it data, and configures the learning parameters, but the final product—the code—is generated independently by the system, is the developer the owner? Is it the company behind the data or the platform that trained the model?

This is not a hypothetical scenario. It’s playing out in courtrooms, patent offices, and legal think tanks around the world.

Understanding the Types of AI Creations

To unpack the issue, it helps to distinguish between different types of AI-driven work:

  • AI-Assisted Creation: A human uses AI tools as support (e.g., using AI to generate image suggestions for a design). Here, IP rights usually stay with the human.
  • AI-Generated Creation: The final product is produced entirely or mostly by AI, without detailed human direction. This is the grayest area.
  • Autonomously Invented Algorithms: The AI system is responsible for developing new algorithms or processes, such as optimizing supply chain routes or discovering new mathematical formulas.

Each of these scenarios raises unique legal and ethical questions. But they all boil down to the same dilemma: should a machine be recognized as an inventor or author?

What the Law Says (and Doesn’t Say)

In the U.S., the Patent and Trademark Office (USPTO) and the Copyright Office have taken a firm stance: only natural persons (i.e., humans) can hold copyrights or patents. This means that any submission must identify a human as the inventor or author, even if the AI was the actual creator.

Other countries are starting to diverge. The United Kingdom and Australia have seen cases where AI-generated inventions were debated in court. In a notable instance, Dr. Stephen Thaler submitted patents listing his AI, DABUS, as the sole inventor. Courts in the U.S. and UK rejected the claims, while Australia briefly accepted them before backtracking.

These mixed responses reveal how ill-equipped current legal systems are for this technological reality.

Corporate Ownership and the Role of Data

The question of ownership becomes even murkier when you consider the data used to train the algorithm. AI systems are only as good as the data they’re fed—often vast, proprietary sets collected over years.

If Company A develops the AI platform, and Company B licenses it to generate new IP, who owns the result? The answer often comes down to contract law rather than IP law. It’s increasingly common for companies to bake IP clauses into licensing and partnership agreements.

Moreover, data privacy and ownership further complicate the conversation. If an AI model is trained on user-generated data, do those users have any rights over the model’s outputs? So far, most jurisdictions say no, but that could change.

What Startups and Innovators Should Do

For entrepreneurs working in AI or using AI to develop products, these are not distant academic concerns—they’re core business risks. Here are some ways to navigate this tricky terrain:

  • Document Human Contribution: Make sure there’s a clear record of how humans were involved in shaping, guiding, or supervising the AI’s output.
  • Review Licensing Agreements Carefully: If you’re using third-party AI tools, check who owns what under the hood.
  • File IP Early: Even provisional patents can help stake a claim to ownership before a competitor beats you to it.
  • Consult with an IP Attorney: Especially one with experience in AI or emerging technologies.

A Glimpse at the Future

Ultimately, the law will need to evolve. There is growing recognition that traditional IP frameworks are too rigid to handle AI’s capabilities. Some experts advocate for a new category of IP ownership—something between traditional authorship and corporate control.

Others suggest updating definitions of “inventor” or “author” to allow for shared credit between AI and human operators. Whether this happens soon or decades from now will depend on political will, judicial interpretation, and economic pressure.

What’s clear is that the future of innovation is entangled with AI. If we don’t adapt our IP systems, we risk stifling the very innovation these systems were designed to protect.