CCI staff share recent surveys, reports and analysis on risk, compliance, governance, infosec and leadership issues. Share details of your survey with us: editor@corporatecomplianceinsights.com.
Life sciences CCOs report highest compensation in survey
Compensation for chief compliance officers at organizations of all types continues to rise, according to a survey by executive recruiting firm BarkerGilmore, though CCOs in certain industries are likely to be much more well-compensated.
Among CCOs in life sciences, the median annual compensation was $785,000, while those in the consumer sector had a median compensation amount of $750,000, according to the survey of more than 250 CCOs. Nonprofit CCOs posted the lowest median compensation, $250,000. Similar patterns showed up in salary increase over 2025: the median salary rose by 5% in life sciences but just 2% in nonprofits.
Company revenue is also a driving factor, the report found. At the 90th percentile in total compensation, for firms both public and private with revenues over $5 billion, total compensation for CCOs, including base salary, long-term incentives and bonuses, exceeded $1.2 million annually, with those at private firms earning slightly more. Among companies with revenues of less than $500 million, the 90th percentile total compensation for CCOs at public companies was $650,000 and $710,000 at private ones.
Other key findings:
- Nearly two-thirds (62%) of CCOs in the survey said they were unlikely to change employers.
- Among those who are considering a job change, 19% said better compensation was the most motivating factor, followed by desire for new challenges or opportunities (16%) and stronger leadership alignment (13%).
- Women were better compensated than men, with the average woman (women accounted for 38% of the survey population) reporting total compensation of $530,000 and the average man reporting total compensation of $495,000.
Claude, ChatGPT score well on set of legal tasks
A recent study by Percipent, a legal services provider, tested how well popular public AI models, such as ChatGPT, Claude and Gemini, performed a variety of legal tasks, finding that several models performed well at certain tasks but struggled when it came to more complicated ones.
For an insurance coverage task, Percipent asked AI to analyze a duty-to-defend scenario under Illinois law and produce a coverage memo with specific citations. In employment law, Percipent asked AI to analyze a former employee’s potential discrimination claims and recommend a motion-stage posture and a settlement strategy.
The AI outputs were graded on a 100-point scale by practice-specific attorneys with an average of more than 25 years of experience. AI models used in the study were versions of Claude by Anthropic, versions of ChatGPT by OpenAI, Gemini by Google, Grok by xAI, versions of Kimi by Moonshot AI and DeepSeek AI.
Claude models scored the highest among the four tasks — insurance coverage, employment law, litigation document review and contract review and redline — though the survey’s authors cautioned against interpreting their results as having declared a winner. Claude tied with a ChatGPT model on employment law but was behind Claude models on the others.
“The point of the work is methodological,” the study said, “to show what a defensible legal AI evaluation looks like, to provide a repeatable framework for measuring real legal performance and to give legal teams a basis for asking sharper questions when they evaluate a tool.”
Routine work revealed near-parity across most of the models, with nine of the 10 variants clustered within eight points of each other on document review, but complex reasoning revealed substantial performance differences. On the insurance coverage analysis task, for example, models were separated by as much as 37 points.
Chad Main, attorney and founder of Percipient, told CCI that the extent to which teams will adopt AI for certain legal tasks depends in part on the risk posture of the organization in question.
“You can look at the stuff (AI models) did really well and go, ‘Look, if perfect is not required of this particular piece of legal work, hey, here we go, it’s a good example,’” Main said. “But then you go back to the coverage and employment, where it scored lower. The people that are risk-averse, especially coverage, you can say, ‘Hey, look, we’ve got to analyze policies, you’ve got to analyze facts, you’ve got to analyze human behavior,’ and no AI can do that all to the point of a human at this point.”
Other key findings:
- Reasoning-focused models consistently led analytical tasks. Across insurance coverage, employment law and contract review benchmarks, extended-thinking and reasoning-mode models outperformed their standard counterparts.
- Thinking longer matters when legal analysis is difficult. While reasoning models showed only modest gains on document review, they delivered significantly stronger performance on tasks requiring multi-step legal analysis and application of authority.
- Noise resistance is a critical differentiator. Models varied substantially in their ability to ignore irrelevant facts, avoid hallucinated authority, and focus on legally responsive issues, capabilities that directly impact reliability in real-world legal workflows.
A third of corporate clients considering breaking up with firms over AI
A massive gap splits corporate clients’ AI wants and what they’re getting from firms they work with, a new survey from Thomson Reuters reports.
More than three-fourths (78%) of respondents said receiving AI-enabled quality improvements from the firms they work with is very important or essential. However, only 6% said most or all of their providers deliver these improvements in AI.
That leaves 31% reconsidering relationships in the next 12 months with firms that aren’t delivering on AI with a third of them estimating that more than $1 million in annual work is at risk.
The findings were based on a survey of 1,816 professionals across law, tax, audit, accounting, compliance, risk and global trade. Respondents span private practice firms as well as in-house corporate and government departments across 62 countries.
Just having a named AI strategy makes the technology work better for organizations, according to the survey. In firms and departments with a named strategy, 66% of professionals say AI is meeting or exceeding expectations for creating value at work. Where there is no active strategy, that figure drops to 22%, the survey said.
For organizations adverse to AI or slow to adopt the technology, professional consequences can be severe, the report warns.
More than a third (34%) of professionals use shadow AI, or tools their organization hasn’t sanctioned, in ways it can’t see, “a sign that adoption is outpacing governance, and a quiet liability for the organizations where it’s happening,” the report’s authors said.
60% of healthcare leaders checking off inadequate compliance
Healthcare leaders are checking off on compliance despite internal audit findings, according to a survey by Omega Systems, which found 60% of healthcare leaders have self-attested to HIPAA compliance while knowing their own audits flagged unresolved vulnerabilities and only 33% never attest until all risks are remediated.
Omega Systems surveyed 200 healthcare executives, IT leaders and practice administrators in the US.
With an upcoming change to the HIPAA security rule looming, nearly one in three (28%) respondents say they can’t meet the update’s requirement for written 72-hour data recovery procedures.
The survey also asked directly about cybersecurity and third-party vendors and found 85% of healthcare practices experienced at least one operational disruption caused by a third party.
