Industry executives and experts share their predictions for 2025. Read them in this 17th annual VMblog.com series exclusive.
By Greg Whalen, CTO at Prove AI
AI has swiftly evolved from an emerging technology to a critical
priority for businesses. In the wake of tools like ChatGPT, 85% of enterprises
now rank AI among their top five priorities, and 97% of business
leaders report a growing urgency to adopt it. However, while enthusiasm for AI
is high, it remains in the early stages of its lifecycle, with significant
challenges to overcome – particularly in managing risk.
One of the most pressing issues facing organizations today is AI
risk mitigation. As companies move AI from experimental pilots to production,
the benefits and risks of adoption are becoming increasingly clear. Prove AI’s recent report, The Essential Role of Governance in Mitigating AI Risk, highlights
that only 5% of organizations have implemented an AI governance framework, with
82% of executives saying implementing AI governance solutions is a somewhat or
extremely pressing priority.
In 2025, governance will be a critical enabler of AI innovation,
helping businesses navigate challenges while unlocking the full potential of
their investments. Based on industry trends and conversations with business
leaders, here are three governance developments I expect to define the year
ahead:
1. Increasing
Delegation of Action to AI Will Drive an Urgent Need for Governance
AI is evolving rapidly, with autonomous agents capable of
executing complex tasks becoming a focal point
of innovation. For example, an AI agent could be used to
provide 24/7 customer support and take actions to resolve customer inquiries.
However, with autonomy comes risk. As these systems begin to
make independent decisions, organizations must address questions of
accountability, transparency, and reliability. What data informed the decision?
How did the system reach its conclusion? Governance frameworks will play a
pivotal role in addressing these challenges by ensuring data quality,
monitoring system behavior, and providing explainability for AI-driven
outcomes.
In 2025, the risk of agentic AI will intensify the focus on
governance, with businesses prioritizing tools and strategies that enable
transparency and responsible deployment of these advanced systems.
2. Continuous,
Automated Governance Will Become Essential
Governance in AI is poised to follow the same evolutionary path
as past technologies like SaaS and Web2. Initially, organizations took an ad
hoc approach, creating manual processes to meet specific needs. Over time,
these processes give way to automation, enabling greater scalability and
efficiency.
With businesses increasingly clear on their AI use cases and the
necessary controls, automated governance tools will emerge as a natural next
step. These systems will allow organizations to track how AI interacts with
data, how datasets influence AI behavior, and how multiple stakeholders
contribute to the lifecycle of an AI system.
The shift to automation will also enable companies to adapt more
readily to evolving compliance and reporting requirements, making governance
more agile and less resource-intensive. In the long term, automated governance
will be crucial for organizations aiming to scale their AI efforts responsible
and efficiently.
3. Building Trust
Through Transparent AI Ecosystems
As the AI landscape has grown increasingly competitive and the
pool of companies developing foundation models has shrunk, more businesses are
building third-party applications on top of LLMs to license to third parties or
use for internal operations. For reference, the AI SaaS market was valued at
$71.54B in 2024 and is projected to reach
$775.44B by 2031.
With this trend, more organizations are becoming involved at
various stages of AI development, with each party largely
responsible for the workloads they develop. This makes it crucial to
establish a clear chain of trust between multiple parties to ensure regulatory
compliance, transparently validate AI data, and improve explainability
– key factors for earning trust. Distributed ledger technology (DLT) has
already seen significant adoption and proven
reliability for these chain of trust tasks in highly regulated industries like
financial services, supply chain logistics, and food safety, given the fact the
technology facilitates multi-party access and records data in a time-stamped
and tamper-proof nature. For example, a payment processing company operating
across multiple countries can leverage DLT to demonstrate their use of consumer
data complies with regional laws and regulations quicker and more provably than
possible with traditional audits and artifact submission.
In 2025, I expect more organizations to adopt transparent,
collaborative approaches to governance, using established tools and
methodologies to build trust into their AI ecosystems.
Preparing for 2025
As businesses continue to embrace AI, governance will no longer
be optional – it will be essential for managing risk, maintaining trust, and
achieving sustainable growth. From the risk of autonomous AI agents to the
adoption of automated governance tools and the focus on trust-based ecosystems,
the trends shaping 2025 reflect a broader recognition of governance as a
cornerstone of AI success.
Next year will be pivotal for organizations striving to
integrate AI responsibly. Building out event collection capabilities that
ensure data observability for both internal and external audits will prepare
companies for success, regardless of what direction AI takes, and those who
prioritize governance now will position themselves to unlock the full potential
of this transformative technology.
##
ABOUT THE AUTHOR
Greg
Whalen is CTO at Prove AI, bringing a wealth of experience scaling startup and
enterprise SaaS businesses. Greg is focused on leading our product development
team through launch and successful adoption of our AI Governance application,
Prove AI.
Greg
has launched and scaled multiple startups and new enterprise product lines. His
teams have invented and operationalized efficient AI/ML components as part of
these solutions across marketing, customer support, health care, and industrial
sectors. Through his work applying AI to regulated use cases, he’s passionate
about finding efficient, pragmatic solutions to emerging needs of AI
Governance. Under his leadership, his high-availability products have
consistently grown revenue 5x-10x while exceeding desired service levels.
Greg
has a strong technical background and global leadership in product management,
software development, and M&A. In the past, he has served as CTO for
Xendit, a southeast Asia payments unicorn, GM for Amazon WorkMail, AWS’s public
email messaging service, the Asia, Pacific and Japan director for Pivotal’s
Data Science business, and product/engineering leader (VP/CTO) for Experian
CheeahMail and Derm101.com. Prior to his work in software, he worked in AI/ML
research as a PhD candidate in Columbia University’s Natural Language
Processing group, where he published research on the use of machine learning
for automated briefings of cardiothoracic surgeries. He holds a MS in Computer
Science from Columbia University and a Bachelor of Arts in Computer Science and
Music from Wesleyan University in Middletown, CT.
