Generative AI: the new business imperative

Generative artificial intelligence (GenAI) is going gangbusters. Since ChatGPT broke into the mainstream in late 2022, GenAI has rapidly evolved from a curiosity into a business game-changer. Companies of all kinds are investing accordingly:

• 96% of the world’s companies are now actively channeling resources into developing or deploying GenAI tools
• Machine learning engineers now command $900,000 salaries
• Venture firms are writing checks worth $15 billion to GenAI startups

New opportunity for innovation and productivity

From machine-generated poetry to revolutionary innovations in semiconductor design, GenAI has the potential to disrupt, streamline, and augment virtually every business function across countless different industries.

Part of GenAI’s power lies in its inherent ease of adoption: with prompt-based, natural-language interfaces, GenAI is becoming accessible to millions of new users who were previously shut out of first-gen AI tools. 

GenAI has the potential to automate away the busywork that currently consumes the attention of human employees. McKinsey estimates that GenAI tools could ultimately free up 60% to 70% of employees’ time.

The combination of a remarkably low barrier to entry and incredible potential to drive value across countless domains make GenAI a powerful force.

‍

GenAI makes data management harder 

However: there’s another side to the story. Even as companies are investing time, money, and resources in GenAI, three-quarters of global businesses — including tech giants such as Samsung and Apple — have banned or are considering banning their employees from using GenAI tools. 

What’s leading so many companies to floor the gas pedal while simultaneously slamming on the brakes? GenAI radically changes companies’ relationship with both consumer and corporate data. 

GenAI’s insatiable appetite for data 

GenAI’s power comes at a price: these transformative technologies have an insatiable appetite for data. 

Large language models (LLMs) and other GenAI tools work by detecting statistical patterns in vast pools of training data: OpenAI’s GPT-4 model, for instance, was originally trained using billions of words of text and hundreds of millions of images distilled from a 45-terabyte data dump of the entire internet. Huge amounts of public data — some of it subject to privacy or copyright laws — is baked into the foundation models that underpin virtually all GenAI tools. 

Typically, GenAI models need to be re-trained or fine-tuned to ensure they produce appropriate outputs for a given use-case, then further refined and improved once they’re deployed. That requires companies to feed large amounts of additional consumer and corporate data into their AI models.

Business struggle with data activation already 

Acquiring and activating data is already a critical business imperative. But AI makes it even more important: companies will need vast quantities of data to train, fine-tune, and continually refine AI models. If companies lack access to high-quality data, their AI projects are dead on arrival.

Companies already struggle to activate their most valuable data: at present, according to IDC, 44% of business data isn’t effectively captured, and 43% of captured data goes unused. The result: less than a third of the data flowing through today’s organizations is actively utilized to drive value for businesses and their customers. 

GenAI tools will change this. As AI technologies grow more commonplace, businesses will increasingly find that all their data — both consumer and corporate — will touch an AI model at some point in its lifecycle. Already, 40% of organizations have thousands of AI models deployed across their operations; that number will only increase in coming months and years.

Employee usage risk

Once a GenAI tool is deployed, it continues to absorb data as it’s leveraged by end-users. It’s incredibly easy, for instance, for employees to copy-paste sensitive data into AI prompts; among organizations that have faced AI security or privacy incidents, 60% report that their data was compromised by the actions of their own employees. 

As GenAI tools grow more commonplace, employees will use them whether their employers like it or not. In fact, Gartner estimates, by 2026 at least 5% of workers covered by GenAI bans will ignore the prohibitions and use the tools to simplify their daily workflows.

The reality is that GenAI bans won’t deliver real data security — but they will stifle innovation and make it harder for companies to keep up with the competition. We need a smarter approach that makes it possible to maintain data privacy and security while still using GenAI to boost productivity and drive business growth.

‍

Making the case for privacy-safe AI 

Combine these challenges with the rapid evolution of both consumer expectations and the global regulatory landscape, and the complexity of the risks facing today’s businesses becomes clear. GenAI makes data more important than ever — but it also makes managing data responsibly, and securing continuing access to data, more difficult than ever before. 

As businesses, we need to reconcile these two opposing faces of GenAI: 

1. Take advantage of transformative new growth opportunities. 
2. Manage a broad range of emerging data protection, IP leakage, regulatory, and reputational risks that collectively pose an existential threat to new AI initiatives. 

This requires true privacy-safe infrastructure — giving both consumers and companies the agency and data security they deserve.

By 2028, it’s estimated that over half of companies now building custom AI models will have shelved the projects due to spiraling costs and ever-increasing complexity. Even today, just 54% of enterprise AI projects make it from pilot to production. Data security and privacy are key factors in determining which organizations succeed and which struggle.

At Ketch, we believe that this calls for a new paradigm: privacy-safe AI, grounded in automated technologies that ensure end-to-end data data privacy across the entire AI value chain. 

‍

Ketch Data and AI Governance: privacy-safe data in a GenAI world 

Companies that build out privacy-safe infrastructure gain the ability to aggressively acquire, retain, and utilize data — secure in the knowledge that they’re responsibly managing both their customers’ sensitive information and their own proprietary IP and business data.

At Ketch, we’re leveraging our extensive experience developing programmatic privacy solutions to help organizations develop the privacy-safe AI solutions needed to win in today’s rapidly evolving data economy. Using automated permissioning and data control technologies, we’re helping organizations to innovate their data infrastructure, ensure regulatory compliance, and responsibly mobilize data to drive sustainable growth in the global AI economy. 

We’re building three critical solutions to help businesses unlock AI value without getting stalled by privacy and compliance gaps. 

1. Ketch GenAI Shield: protect business confidential data in AI initiatives

In the AI era, you can’t afford to simply hope that your employees don’t feed sensitive corporate or consumer data into your AI models. You need to know that your data is only being used in safe and appropriate ways. Ketch GenAI Shield helps you protect against inadvertent misuse of business confidential data in AI initiatives.

• Use Enterprise Data Fortification™ to protect company confidential and sensitive data against leakage in open source AI workflows
• Set policy for which data types can be shared with AI models, flagging misuse or non-sanctioned data use
• Understand the types of data you have within your data systems, assigning custom data classifications that are relevant and meaningful to your business

2. Ketch Permissioning Engine for AI: enable privacy-safe data for AI models

In the GenAI era, consumers will only grant access to their data to organizations that respect their agency and their privacy choices. ‍Ketch Permissioning Engine for AI ensures only privacy-safe, permissioned data enters AI models. Enable your teams to unlock the value, productivity, and competitive advantages of Generative AI.

• Collect meaningful consent from consumers anchored in the specific purposes for which data is used
• Fulfill data subject requests in AI models
• Get purpose-built, sustained and responsive compliance with current and future privacy and AI laws 

3. Ketch AI Model Governance: understand and manage your AI footprint

With regulators rushing to adapt to the AI era, organizations need to be able to set clear central policies defining how data can and can’t be used across their AI ecosystem. Ketch AI Model Governance is a system of record for AI governance. 

• A global system of record for AI models: recording each model’s intention, attestations of non-bias, target audience and other business or legally mandated obligations in a centralized model registry
• Track and audit AI models, understanding the data inputs and outputs across each AI initiative
• Gain a new level of insight and risk mitigation into your data and AI ecosystem, with intelligence for managing risks and making data-driven decisions for compliance

Together, these capabilities lie at the heart of privacy-safe GenAI innovation. They make it possible to drive business growth while providing dependable security and real privacy protections for both consumer and corporate data. 

‍

Data privacy is the barrier to overcome for AI success

By 2026, organizations that successfully build out privacy-safe AI solutions grounded in transparency, trust, and data security will realize a 50% gain across key metrics including user acceptance, technology adoption, and total business value. Companies that fall short on those measures, of course, will be left in the dust.

A marathon, not a sprint

At Ketch, we know the AI arms race isn’t a sprint — it’s a marathon. The goal isn’t simply to surge ahead at any cost. It’s to build a stable foundation that ensures permissioned and privacy-safe data flows through all your AI operations, and that your AI projects aren’t merely a flash in the pan.

The bottom line: privacy-safe data is the key to unlocking lasting value for your customers, and lasting growth for your business. In a world that’s being redefined by GenAI, putting privacy and security at the heart of your AI projects is the key to building competitive advantage and finding enduring success.

Learn more about our privacy-safe AI solutions → Request early access. 

‍