As organizations look to embrace a new generation of autonomous AI agents which can access enterprise systems, make decisions and interact with other AI agents, new questions are emerging around how these systems are governed and secured.

As the governance of agentic AI and non-human identities rapidly one of the most significant challenges organizations will facein the coming years, we spoke to Umashankar Lakshmipathy, EVP and Head of Cloud, Infrastructure, and Security Services, EMEA, Infosys, to find out more.

• As organisations move from AI assistants to autonomous AI agents, what new security and governance challenges are emerging?

The move towards autonomous AI agents creates new security and governance challenges because these systems can make decisions and take actions with limited human involvement. Unlike AI assistants that primarily generate information or recommendations, agents can access enterprise data, use applications and tools, communicate with other agents and act on behalf of people or business processes.

This greater autonomy increases the potential business impact of an error, manipulation, or security compromise. An agent may have a valid identity and legitimate system access but could still take action that falls outside its intended purpose. The risk becomes more significant when agents operate across several systems, as one incorrect decision can trigger a sequence of actions that are difficult to detect, explain or reverse.

Organisations must therefore govern more than the AI model. Every agent needs a defined business purpose, an accountable owner and clear limits on the data, systems, and actions available to it. Sensitive or high-impact activities require stronger safeguards, while organisations need sufficient visibility to trace an outcome back to the person or process that initiated it, the agent that acted and the tools it used.

This represents an important evolution in AI governance. Organisations must continue to manage the quality and safety of AI-generated content, but they must now also control the decisions and actions that autonomous agents can execute. Getting that balance right will be essential to achieving the productivity benefits of agentic AI without creating unacceptable operational, financial, or regulatory risks.

• As agentic AI adoption grows, why are “non-human identities” becoming such an important issue for enterprises? What are the risks and opportunities when AI agents begin communicating and making decisions with other AI agents?

Non-human identities are becoming an important enterprise issue because a growing proportion of digital activity is performed by applications, APIs, service accounts, automation and AI agents rather than by employees. AI agents increase the significance of these identities because they can interpret information, make decisions and initiate actions, rather than simply execute predefined instructions.

This makes visibility, ownership, and access management essential. Organisations need an accurate inventory of the non-human identities operating across their environment, together with a clear understanding of their business purpose, accountable owner and access to systems and data. Without this foundation, excessive permissions, long-lived credentials, and identities that remain active after their original purpose has ended create material security exposure.

Communication between agents creates significant business opportunities. Specialised agents can work together to automate complex processes across customer service, operations, software engineering and cyber defense. A process that previously required several teams and systems could potentially be coordinated by a group of agents operating at much greater speed.

However, agent-to-agent collaboration also creates chains of dependency and trust. A compromised or poorly governed agent could pass inaccurate information through a workflow, influence another agent’s decisions or attempt to use a more privileged agent to perform an action that it could not execute directly. These risks can multiply when organisations cannot establish which agent initiated an action or under whose authority it was performed.

Organisations should therefore enable agent collaboration within defined business and security boundaries. The identity, authority and purpose of each agent should be evaluated independently rather than allowing trust or access rights to pass automatically from one agent to another. This approach allows enterprises to benefit from multi-agent automation while preserving control and accountability across the complete business process.

• How can organisations strike the right balance between giving AI agents autonomy and maintaining appropriate oversight and control?

The appropriate level of autonomy for an AI agent should be determined by the business impact of its activities, the sensitivity of the information involved and how easily its actions can be reversed. Applying the same level of oversight to every agent would either create unnecessary friction or leave higher-risk activities insufficiently controlled. For example, an agent that analyses public information or produces an internal summary can operate with relatively light oversight. An agent that accesses customer information, changes production infrastructure, makes regulatory decisions or initiates financial transactions requires much tighter boundaries.

However, organisations should introduce autonomy progressively. Some agents will remain assistive and provide recommendations for people to consider. Others may prepare actions that require human approval. More mature agents can operate independently within approved limits, escalating situations that fall outside their normal authority. Agents performing highly predictable and well-controlled activities may eventually be given broader autonomy, provided their actions remain visible, governed and capable of being stopped.

Human oversight remains important, but requiring approval for every action would undermine the efficiency that agentic AI is intended to deliver. It can also lead to approval fatigue, where reviews become routine and provide little meaningful protection. Human involvement is most valuable for actions that are financially significant, legally sensitive, difficult to reverse or outside the agent’s established operating boundaries.

The aim is not to give agents the maximum possible autonomy. It is to give each agent enough autonomy to deliver business value while keeping its decisions within an acceptable level of risk. Routine and reversible activities can be automated, while consequential decisions remain subject to stronger safeguards and clear organisational accountability.

• Are existing cybersecurity, identity and governance frameworks sufficient for the age of agentic AI, or do organisations need a new approach?

Existing cybersecurity, identity and governance frameworks provide a strong foundation for agentic AI, but they are not sufficient without further adaptation. Organisations do not need to replace established capabilities such as Zero Trust, identity governance, privileged access management and AI risk management. They do, however, need to apply these capabilities more dynamically to agents that can make decisions and take actions at machine speed.

Many traditional controls were designed around granting access to a person, application or service and reviewing that access periodically. Agentic AI introduces a more dynamic operating model. An agent may access several data sources, invoke multiple tools, interact with other agents and make numerous decisions during a single business process. Authentication alone cannot establish whether every subsequent action is appropriate.

Existing frameworks therefore need to be extended in several practical ways. Every agent should have a unique identity, an approved business purpose, an accountable owner and a managed lifecycle. Its authority should be limited to the task it is performing and should not remain active for longer than necessary. Stronger safeguards should apply when an agent accesses sensitive information, uses privileged tools or attempts an action with a significant business impact.

Organisations must also be able to reconstruct how an outcome was produced. This means maintaining a clear record linking the originating person or business process to the agents involved, the data and tools they used, the actions they took and the resulting outcome. This visibility is essential for oversight, investigation, regulatory assurance and rapid intervention when something goes wrong.

The required approach is therefore an evolution of existing security and governance rather than the creation of a separate control environment for AI agents. By extending established frameworks to govern agent identity, authority, and actions continuously, organisations can scale agentic AI while maintaining coherent enterprise-wide control.

Over the last few years, Zoom has probably had one of the most intriguing journeys in the business world.

After booming during the pandemic, the company has since expanded its offerings to cover a huge range of enterprise needs, with contact center services alongside traditional video conferencing and expanding AI tools.

As the company looks ahead to its new AI-enabled evolution, I went to the Zoom Experience Centre in London to find out more.

Refusal to compromise

“When we look at EMEA, we think the landscape is shifting in how organizations think about the technology they rely on,” Louise Newbury-Smith, Zoom’s Head of UKI, explains.

“For years, the pattern has been pretty similar, an organization would adopt a platform and then have to manipulate their own way of working around that platform - we really want to reverse that.”

“There is still some of that friction…(but) what’s changing is that organizations are starting to refuse to compromise - they want platforms to work the way they want to work.”

Zoom opened its UK data center in November 2025, greatly improving the situation for organizations in the region, as data residency choices which didn't exist before are now in place, meaning the likes of local councils, NHS trusts and financial services firms can store their data locally.

Newbury-Smith notes that the EMEA region is not a secondary market for Zoom, but high growth opportunity, a point echoed by Nick Tidd, Head of Global Channel GTM, who said the region offers, “a truly remarkable opportunity for our customers and for our partners, when you look across the product portfolio.”

This opportunity is hooked onto AI, particularly the growing importance of the context layer, a situation where Zoom is ideally placed, Tidd says.

“Think of how Zoom is positioned across the product portfolio - capturing all of those conversations…(and) putting them into the context to capture them in the workflow is where we are very different to others.”

This move was expanded by Leo Boulton, Zoom’s Head of Product, Solution and Industry Marketing, who ran us through the company’s recent releases and product announcements, highlighting the importance of context in Zoom’s concept of “conversation to completion” in its new Unified Context Layer.

“Conversations come natural to us - the way that we work is through conversations, it’s just a natural part of how humans are,” he noted, “so getting the job done should also be a natural thing - not something that requires a change in behavior, but something that should naturally just be part of our everyday work.”

If this doesn’t happen, workers will resist change, Boulton notes - particularly when it comes to AI, which is why Zoom is looking to embed intelligence into work itself.

Context has become a critical part of improving AI performance and usefulness in recent months, and Zoom wants to ensure its tools are as up to date as possible - but this requires some key elements.

“Before AI can truly understand how people work and operate, we have to guarantee it has the highest-quality raw materials,” Boulton points out. “The conversation must be crisp, clear and understood with genuine eye contact – because only with the best ingredients can AI deliver the most value.”

Zoom has been working hard to improve the quality and fidelity of its audio and video for calls and meetings, including the upcoming launch of a real-time “voice smoothing” tool known as “touch up my voice” and a pose correction tool enabling for greater eye-to-eye contact to boost interpersonal connections.

The better your agents understand your business, the more effectively they can work for you, Boulton notes, as he demonstrated how users will soon be able to benefit from a live assist feature which offers proactive suggestions during a call, such as telling you to speak slower or faster.

Zoom Chat is also getting a significant upgrade, making it what Boulton said should be a shared space where people and agents will interact and work together.

“We don’t expect our customers to live entirely inside Zoom,” he said. “True adoption only happens when tools feel natural, which is why our context layer is open by design – bringing the intelligence directly into whatever interface people already choose to work in.”

So it’s clear that Zoom wants to be so much more than your classic video conferencing and chat platform - and its ability to get the most out of your information could make it a very handy tech partner going forward.

“The direction is super clear,” Newbury-Smith concludes, “we're building and serving our customers where they are.”

As workers look for more ways to be productive at home and in the office, wearable technology has often proved a useful tool, offering quite literally a new dimension for users.

Augmented Reality (AR) headsets have been around for a few years, with Microsoft, Meta and Apple all among the big players getting involved - but these headsets have often proved clunky and overbearing.

AR glasses have instead offered a new opportunity for those looking for some extra assistance in everyday life - I went to try out the new Snap Specs ahead of their UK launch.

First impressions

Let’s get the obvious out of the way first - tech-enabled glasses (specifically those made by Meta) are under the microscope somewhat at the current moment in time, with rulings over what they can and can’t do emerging in courts across the world.

Snap has explicitly looked to offer a whole new kind of experience, with the new sixth-generation Snap Specs much more obviously an accessory - it’s pretty obvious when wearing them that (unless you’ve gone full Nathan Barley) these aren’t a typical pair of glasses.

The smooth black material is nice to touch, and sits against your face with no issues. Configuring the glasses takes only a few seconds, with control of the device mainly in AR, with two buttons on the arms of the glasses helping with recording and resetting if needed.

Unlike other AR glasses, the big selling point of Snap Specs is that they don’t need to be tethered to a smartphone, so you can wear them wherever you want - with a portable charging pad helping to top up battery life when needed.

As the Snap Specs are still fairly new, a lot of the app showcases and demos I saw looked to show off how AR can help entertain or educate - but there were some work use cases too.

My glasses came loaded with Google Docs, with typing carried out on a virtual keyboard which could prove perfect if you need to write up some quick thoughts or emails when away from your desk. You might get a few odd looks while typing in mid-air, but when inspiration strikes, needs must - and it can also connect to a Bluetooth keyboard if you’d prefer.

The glasses are also able to capture video clips up to 30 seconds and still images, useful for creative workers across the spectrum - whether you’re getting a picture of a billboard idea, or seeing something in the wild that would be perfect for your next piece of artwork, you can quickly grab an image or video without needing to whip out your phone.

Snap Specs also features a built-in browser, which does everything you would hope - but also doubles as a media viewer, so whether you need to work on or review a new presentation or promotional video, or just want to catch up on some YouTube, the glasses can be a useful tool.

The voice AI assistant was also fast and effective when answering my questions (even if I doubted its accuracy on finding the nearest tube station to our location), and using Google Maps with a heads-up display was an absolute game-changer, although again it might depend on how you feel walking down the street wearing a £2,000 pair of glasses.

Perhaps the most striking demo was the translation tool - when looking at someone speaking in a foreign language, their speech would be instantly translated and shown in a speech bubble overlaying them - essentially like subtitles when watching a TV show. This could be an incredible tool for business meetings in foreign countries, and might well be the killer app for technology such as Snap Specs.

So will the £1,999 Snap Specs be the future of work? Given that the more fun demos I saw involved playing dominoes, improving my football skills, a drawing tutorial and enjoying an interactive Harry Potter experience involving a pet owl and the sorting hat, I’m not sure it’s there just yet - but I can’t wait to see where the technology goes next.

• Study finds people are being more polite to AI
• Some are even apologizing to AI assistants
• Will this really help when the robot uprising comes? Let's hope so

As worries around AI taking over the world continue to rise, new research has shown that most of us are getting increasingly comfortable with our new assistants.

A survey by fonio.ai has claimed Brits are living up to global stereotypes and being more and more polite to AI, even extending that politeness beyond the expected basics.

Along with greetings and thanks, the survey found that a rising number of people are asking how an AI is, much like they would do a real person - and even apologizing when they get something wrong.

Mind your manners

We've already heard about how saying "thank you" can cause incredible amounts of damage to the environment, with OpenAI CEO Sam Altman noting it costs "Tens of millions of dollars" in electricity bills.

The study found 7 in 10 calls included a “please” or “thank you”, while over half (55%) of callers thank the AI at least once.

More than half (52%) also greet their AI assistant with a “hello”, “hi” or “good morning”, suggesting that traditional conversational etiquette carries over to interactions with technology.

The study found around one in eight (12%) apologize to their AI assistant, with 13% saying goodbye before ending the conversation - with a further 7% offering even more of a farewell, saying something like “have a good day” or “take care”.

Slightly more disturbingly, 2% admitted that they even ask the AI assistant how it is, showing how they are treating the interaction much like a conversation with an actual human.

“There’s something very British about thanking a machine for helping you out. Even when people know they’re speaking to an AI, they still bring the same manners they’d use in a conversation with another person," noted Daniel Keinrath, Co-Founder of fonio.ai.

“Whether it’s saying please, apologising or wishing their AI assistant a good day, these little habits show how naturally voice technology is becoming part of everyday life. AI might not need good manners, but that doesn’t mean we’re going to stop using them.”

A February 2026 survey of more than 1,000 people by Future PLC, the company that owns TechRadar Pro, found that around 70% of people are polite to AI when interacting with it, with 12% being polite in case of a robot uprising.