As Canada moves to close its national AI adoption gap, new research from BDO Canada finds nearly half (46%) of Canadian business leaders are experimenting with AI without achieving meaningful ROI.
Only 18% are actively embedding AI into workflows and operations, according to BDO Canada’s AI Vision Report: Past the pilot to the agentic future of work. BDO says the findings point to a second-stage challenge for Canadian organizations: identifying the readiness gaps that limit value, then building the governance, workforce capability and operating discipline needed to scale AI responsibly. For business leaders, this is a movement away from number of pilots, to whether those efforts are improving decisions, managing risk and creating measurable value.
AI Vision Report: Past the pilot to the agentic future of work explores how organizations can move beyond isolated AI pilots toward governed, enterprise-wide adoption. It outlines how leaders can connect AI initiatives to clear business outcomes, accountable decision-making, and measurable value as AI moves from stand-alone productivity tools into more integrated, multi-step workflows.
The report also found that 27% of Canadian business leaders believe AI will have minimal impact on their organization over the next four years–a finding BDO says may point to a visibility gap as AI becomes increasingly embedded into enterprise software, workflows and decision-support systems.
From adoption to responsible scale
BDO says organizations need to scale AI in a way that creates measurable value, manages risk and helps people adapt to new ways of working.
The findings come as businesses begin preparing for agentic AI systems that can support multi-step work, coordinate information across platforms and move teams toward more integrated workflows.
As these capabilities become more common, BDO says organizations will need to treat AI as an operating-model change, not simply a technology deployment. Scaling safely and effectively will require clear governance, ownership, workforce enablement, adoption planning and measurement tied to business outcomes.
According to Gartner, by 2028, one-third of enterprise software applications are expected to include agentic AI capabilities, up from less than 1% in 2024. BDO says this shift will put greater pressure on organizations to build the foundations for responsible scale now, including governance, workforce fluency, workflow integration, and measurement tied to business outcomes.
Informed by BDO’s own AI adoption journey
BDO’s perspective is informed by its own AI adoption journey across its national firm, including the firm’s Client Zero approach to testing, learning, and scaling AI responsibly within its own operations. Through investment in workforce enablement, governed experimentation, workflow integration, and responsible AI practices, BDO has developed practical lessons that inform its work with clients.
Read BDO Canada’s AI Vision Report: Past the pilot to the agentic future of work.
About the report
AI Vision Report: Past the pilot to the agentic future of work is a BDO Canada report examining how Canadian organizations can move from AI experimentation to responsible, enterprise-wide adoption. The report draws on survey of 520 Canadian business leaders who are members of the Angus Reid Forum commissioned by BDO Canada, along with BDO’s experience helping clients connect AI initiatives to business outcomes, governance, workforce readiness and measurable value.
Gaslight malware shows attackers are beginning to target AI-powered security analysis
Posted in Commentary with tags macOS on June 25, 2026 by itnerdThe newly discovered Gaslight malware for macOS highlights an emerging shift in attacker tradecraft: instead of only evading traditional security tools, threat actors are beginning to manipulate AI-assisted analysis itself. By embedding prompt injection techniques designed to mislead or halt LLM-powered malware analysis, attackers are testing how much security teams rely on AI during incident response. As AI becomes more deeply integrated into defensive workflows, organizations will need to treat AI systems as another attack surface, requiring validation, oversight, and resilience against manipulation.
You can find out more details here: New Gaslight macOS Malware Uses Prompt Injection to Disrupt AI-Assisted Analysis
Gidi Cohen, CEO & Co-founder, Bonfy.AI
“Gaslight is a glimpse of where AI‑aware malware is headed—and a reminder that securing the data plane now matters as much as securing endpoints and sandboxes.
This Rust‑based macOS implant doesn’t just steal data and maintain a Telegram‑based C2 channel; it also embeds prompt‑injection content specifically designed to confuse LLM‑assisted analysis pipelines, flooding them with fabricated “system failure” messages to get automated triage to abort or mis‑report. In other words, the malware is actively targeting the AI tools defenders rely on, trying to shape what those systems “see” and how they respond.
For organizations, this means two things. AI‑assisted security workflows need explicit defenses against adversarial content, with clear separation between untrusted artifact data and trusted system messages. And because attackers are gaining more ways to mislead or bypass detection, enterprises must assume that traditional controls will be defeated more often—and ensure they have strong, contextual protection for sensitive data across email, SaaS apps, collaboration tools, and AI systems, so that even when malware slips through or “gaslights” the tools, the blast radius for critical information stays small.”
That should bust any myth that macOS is immune from malware. But realistically, you need to protect every device all the time regardless of OS. Otherwise bad things will happen.
UPDATE: Toghrul Tahirov, Head of AI Governance, Polygraf AI adds this:
“Gaslight is not a sandbox evasion technique. It is a social engineering attack aimed at an AI analyst.
The implant is standard North Korean tradecraft: Telegram C2, Python infostealer, Keychain harvesting. I am specifically amazed that what SentinelOne found is embedded inside it. There are 38 fabricated system messages engineered to convince an LLM-assisted triage agent that its own session is collapsing. They have thought this out! Fake token expiry. Fake OOM kills. Bogus injection warnings. Not to hide from the agent. To make it quit before finishing the job.
We don’t see any architectural separation. A fabricated system message and a real one that look identical to the model. That is not a prompt engineering problem. It is a fundamental design constraint, and adversaries are figuring out how to weaponize it against defenders. And see how fast they are productizing it.
The moment you put an AI agent into your pipeline, that agent becomes a part of your attack surface. Gaslight is the first field sample that treats it explicitly as one.
One can not just handle this sort of issue with a more capable model. Enforcement and security has to happen at the input boundary, kind of stand alone proxy environment, before untrusted content reaches the reasoning layer. That is the problem Polygraf’s AI Behavioral Control Plane addresses.”
Leave a comment »