AI washing: signs, symptoms and solutions for investment stakeholders
Although some firms genuinely apply machine learning (ML) and AI to improve investing, others make claims that do not match reality. These firms may use buzzwords such as 'AI-driven' or 'machine learning-enabled' without truly integrating these tools into their investment processes.
Consequently, clients and investors may be misled into believing they are investing in innovative, cutting-edge strategies when they are not.
This phenomenon is known as 'AI washing' – the act of falsely or overly inflating claims about the use of AI in financial products or services.
The CFA Institute recently published AI Washing: Signs, Symptoms and Suggested Solutions – a report that examines what AI washing is, why firms engage in it, how it affects clients and the broader development of AI, touching on the ethical, regulatory and technical measures that can help address it. It also offers guidance to asset owners on how to spot both genuine AI use and inflated claims in the marketplace.
According to Nvidia's State of AI in Financial Services: 2025 Trends report, 57 per cent of respondents in a global survey of financial professionals are using or considering AI for data analytics, and generative AI usage has risen sharply to 52 per cent from 40 per cent in 2023.
BT in your inbox
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
Sign Up
Sign Up
In addition, 37 per cent report AI-driven operational efficiencies, and 32 per cent believe AI offers a competitive advantage. The use of AI in trading and portfolio optimisation has increased to 38 per cent from 15 per cent, while its application in pricing, risk management and underwriting has grown to 32 per cent from 13 per cent.
Barriers to AI adoption
True AI in finance involves systems that process large data sets, learn patterns and make decisions – such as predicting asset prices or optimising portfolios. These efforts require serious investment in talent, technology and time.
Many investment firms, however, either lack the resources or are unwilling to overhaul their existing processes to meaningfully incorporate AI. Instead, they may add small AI elements (such as using a chatbot or large language model) but advertise their strategy as 'AI-powered', which is deceptive if these tools do not play a central role.
Barriers to real AI adoption in investing are high. Financial data is often messy, sparse and hard to predict. Unlike other industries where data is more abundant and easier to model, investment forecasting requires handling noisy, volatile and complex inputs. Consequently, many firms hesitate to disrupt their existing models that already perform well.
AI washing is particularly dangerous because it undermines explainable AI – a movement focused on making AI systems more transparent, understandable and trustworthy – especially for non-technical users. If firms exaggerate or hide how they use AI, it becomes harder for stakeholders to assess the real value or risks of these tools.
The CFA Institute report asserts that investors deserve transparency about what technologies are being used, how they work and whether they deliver value. Firms should avoid overhyping their use of AI just to attract clients or compete with rivals. Instead, they should be transparent about how they use AI, what it adds to their process and what limitations exist.
Asset managers or asset owners must be able to provide sufficient detail regarding why and how they implement AI technology in their process, what specific frameworks they use, and what results or improvements they observe from using AI.
This recommendation is in line with the ethical principles of transparency and duty to clients as set out in the CFA Institute Code of Ethics and Standards of Professional Conduct.
Asking the right questions to spot AI washing
To help stakeholders – especially asset owners and prospective clients – spot AI washing, the CFA Institute report suggests a range of questions that asset owners and prospective clients can pose to asset managers that claim to use AI.
Some of these questions demand some level of technical familiarity with AI and ML, which speaks to the fact that asset owners themselves must develop some minimal competence in AI methodologies.
Below is a list of pertinent questions for consideration:
Can you specify what type of algorithm or combination of algorithms you are using and how it enhances the forecasting of asset returns?
How does your AI-driven model outperform simpler models? Can you provide a quantitative comparison of relevant performance metrics?
What data sources are you using to train your model(s), and how do these sources integrate with the rest of your process, if at all? Are you using alternative data, such as satellite imagery or sentiment analysis of earnings calls?
What preprocessing and feature selection techniques are used to prepare the raw data for input into your model(s)? Do you use fundamental features, such as earnings surprise, price momentum, or other signals and indicators?
Do you standardise or normalise the input features, and what techniques do you use to handle missing data, outliers and limited data sets?
How do you maximise model interpretability? Is it through model choice or post-implementation communications? If the latter, can you give some concrete examples?
Can you provide an example of a recent investment decision that was influenced by the model's output? How was the rationale for that decision explained to the investment team?
How do you validate the robustness of the models you develop? What precautions do you take to guard against overfitting? For example, how do you tune hyperparameters in your models? How do you monitor model drift, and what mechanisms are in place to retrain the models and/or adapt to shifts in the market landscape?
What governance structures are in place to ensure the responsible use of AI firmwide? Do you have an internal AI audit process, and how often are the models reviewed for compliance with generally accepted standards and protocols?
If you use outsourcing for some or all of your AI technology needs, what processes are in place to ensure the quality and robustness of the services and products used in your investment process?
Transparency is non-negotiable
Firms selling financial products should conform to the same standards of transparency that stakeholders demand from other types of products. This idea applies to the use of AI technology as well.
Unfortunately, because of AI's headline-grabbing popularity, some investment firms may rush to exaggerate their success in applying AI technologies to their investment processes. Such instances of AI washing have increasingly become the subject of heightened scrutiny from the investment community, including regulators.
By understanding and learning to detect AI washing, stakeholders can help minimise and eventually eliminate this phenomenon, resulting in better investment outcomes.
The writer is a senior affiliate researcher with CFA Institute and author of the report AI Washing: Signs, Symptoms and Suggested Solutions. He is currently the co-editor of The Journal of Financial Data Science, on the editorial board of The Journal of Portfolio Management and a member of the board of directors of the Financial Data Professional Institute.
This content has been adapted from an article that first appeared on the CFA Institute Research & Policy Center's website.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Business Times
4 hours ago
- Business Times
Trump to hike India tariffs within 24 hours, eyes pharma and chip duties next
[WASHINGTON] US President Donald Trump said he would raise tariffs on Indian goods 'over the next 24 hours' in response to New Delhi's continued purchases of Russian oil. Trump announced a 25 per cent duty on India's exports to the US and has threatened repeatedly to increase that rate to punish the country for buying Russian energy, an effort to pressure Russian President Vladimir Putin to end the war in Ukraine. 'We settled on 25 per cent but I think I'm going to raise that very substantially over the next 24 hours, because they're buying Russian oil,' Trump said on Tuesday (Aug 5) in a CNBC interview. 'They're fuelling the war machine. And if they're going to do that, then I'm not going to be happy.' Trump has escalated his fight with India over trade, unilaterally imposing a tariff rate after months of negotiations failed to secure a deal. He accused New Delhi of refusing to ease access for American goods and criticising its membership in the Brics group of developing economies. The US president has also set an Aug 8 deadline for Russia to reach a truce with Ukraine, with the administration threatening so-called secondary sanctions on countries that purchase energy from Moscow. Ukraine's allies say those purchases prop up Putin's war effort. Trump in the interview said that if energy prices went down it would undercut Putin's ability to continue his invasion of Ukraine – now in its fourth year. BT in your inbox Start and end each day with the latest news stories and analyses delivered straight to your inbox. Sign Up Sign Up 'If energy goes down low enough, Putin's going to stop killing people,' Trump said. 'If you get energy down another US$10 a barrel, he's going to have no choice, because his economy stinks.' The Indian government has indicated it intends to continue talks with the US in hopes of securing lower tariffs. It has also called Trump's threat over energy purchases unjustified. India is considering ramping up natural gas purchases from the US and increasing imports of communication equipment and gold. In the same interview, Trump also said US tariffs on semiconductor and pharmaceutical imports would be announced 'within the next week or so,' as the administration prepares to target key economic sectors in its effort to remake global trade. 'We'll be putting an initially small tariff on pharmaceuticals, but in one year – one and a half years, maximum – it's going to go to 150 per cent and then it's going to go to 250 per cent because we want pharmaceuticals made in our country,' Trump said. 'We're going to be announcing on semiconductors and chips, which is a separate category,' he added. The Commerce Department has been investigating the semiconductor market since April to set the stage for possible tariffs on an industry that's expected to generate nearly US$700 billion in global sales. Under Trump, the US has already imposed levies on imports of cars and auto parts as well as steel and aluminium. Levies on imported chips threaten to sharply increase costs for large data centre operators including Microsoft, OpenAI, Meta Platforms and that plan to spend billions of US dollars on purchases of advanced semiconductors needed to propel their artificial intelligence businesses. The president has also threatened debilitating tariffs on the drug industry in an effort to force manufacturing back to the US. Trump recently demanded major suppliers of medicines drastically cut costs or face additional, unspecified penalties. The world's largest drugmakers, including Merck and Eli Lilly, operate scores of manufacturing sites across the globe. Nearly 90 per cent of US biotech companies rely on imported components for at least half of their approved products, according to the Biotechnology Innovation Organisation. The sectoral tariffs on pharmaceuticals, metals and other industries stem from trade investigations that can last about nine months and are imposed on national security grounds under Section 232 of the Trade Expansion Act. It's seen as stronger legal footing than the emergency powers Trump used for his country-specific levies, which face court challenges. Those so-called reciprocal tariffs are slated to go into effect on Thursday. BLOOMBERG
Business Times
5 hours ago
- Business Times
US trade deficit hits nearly 2-year low in June; China gap plunges
[WASHINGTON] The US trade deficit narrowed in June on a sharp drop in consumer goods imports, and the trade gap with China shrank to its lowest in more than 21 years, the latest evidence of the imprint on global commerce President Donald Trump is making with sweeping tariffs on imported goods. The overall trade gap narrowed 16 per cent in June to US$60.2 billion, the Commerce Department's Bureau of Economic Analysis said on Tuesday (Aug 5). Days after reporting that the goods trade deficit tumbled 10.8 per cent to its lowest since September 2023, the government said the full deficit including services also was its narrowest since then. Exports of goods and services totalled US$277.3 billion, down from more than US$278 billion in May, while total imports were US$337.5 billion, down from US$350.3 billion. The diminished trade deficit contributed heavily to the rebound in US gross domestic product during the second quarter, reported last week, reversing a drag in the first quarter when imports had surged as consumers and businesses front-loaded purchases to beat the imposition of Trump's tariffs. The economy in the second quarter expanded at a 3 per cent annualised rate after contracting at a 0.5 per cent rate in the first three months of the year, but the headline figure masked underlying indications that activity was weakening. Last week Trump, ahead of a self-imposed deadline of Aug 1, issued a barrage of notices informing scores of trading partners of higher import taxes set to be imposed on their goods exports to the US. BT in your inbox Start and end each day with the latest news stories and analyses delivered straight to your inbox. Sign Up Sign Up With tariff rates ranging from 10 to 41 per cent on imports to the US set to kick in on Aug 7, the Budget Lab at Yale now estimates the average overall US tariff rate has shot up to 18.3 per cent, the highest since 1934, from between 2 per cent and 3 per cent before Trump returned to the White House in January. China trade gap A centrepiece of Tuesday's report was the latest steep drop in the US trade deficit with China, which tumbled by roughly a third to US$9.5 billion in June to its narrowest since February 2004. Over five consecutive months of declines, it has narrowed by US$22.2 billion – a 70 per cent reduction. US and China trade negotiators met last week in Sweden in the latest round of engagement over the trade war that has intensified since Trump's return. The US currently imposes a 30 per cent tariff on most Chinese imports, which has fuelled a steep drop off in inbound goods traffic from China. Imports from China dropped to US$18.9 billion, the lowest since 2009. The trade negotiators have recommended that Trump extend an Aug 12 deadline for the current tariff rate to expire and snap back to more than 100 per cent, where it had briefly been earlier this year after a round of tit-for-tat increases by both sides. 'We're getting very close to a deal,' Trump said on Tuesday in an interview on CNBC. 'We're getting along with China very well.' REUTERS
Business Times
8 hours ago
- Business Times
No safe assets over the long run – but some lose less than others
IMAGINE receiving $1 million today, along with the responsibility to safeguard and grow it over the next decade. Your primary goal: preserve its real value – and ideally increase it. How would you allocate this capital? There is no straightforward answer. History shows that even the most secure-seeming options carry hidden risks. Capital at risk, always US financial data from 1900 to 2024 shows that inflation averaged 3 per cent a year. This means over a century, one dollar eroded to less than four cents – a loss of more than 96 per cent in purchasing power. What if you put your cash in a savings account? That gives you interest and feels safer. Over the past century, savings accounts in countries like the US and other western nations have on average kept pace with inflation. Short-term saving rates, proxied by short-term US T-bills, averaged 3 per cent a year. Averages mask significant losses, however. Financial repression in the 1940s and early 1950s saw interest rates being held artificially low while prices crept higher. This was done to lower the debt of governments heavily indebted after World War II. Savers suffered a real loss in purchasing power of more than 40 per cent. As at 2025, a new era of financial repression appears to be underway. The inflation spike of 2022, combined with interest rates lagging behind, caused a real loss in value of nearly 20 per cent. Savers are still down about 10 per cent relative to 2010 levels. With real interest rates near zero in 2025, catching up will be difficult. BT in your inbox Start and end each day with the latest news stories and analyses delivered straight to your inbox. Sign Up Sign Up These episodes underscore a fundamental truth: even assets that feel safe – like savings accounts – can expose investors to real, lasting losses. That brings us to a broader point – capital is always at risk. Whether you choose to save or invest, you're making a bet. Inflation and market volatility are ever-present. Government bonds: Safer, but safe enough? For many investors, the next step beyond saving is government bonds. They typically offer about 1 per cent more yield than a savings account and are often viewed as a safer alternative to equities. But safe from what? Bond investors have faced challenging periods since 1900. After World War I, a post-war economic boom led to rising inflation, which eroded the purchasing power of government bonds issued during the war. A similar pattern followed World War II – artificially low interest rates and a prolonged bond bear market. Then came the 'bond winter' of the 1970s, when bondholders lost nearly 50 per cent in real terms. That's not just volatility, that's wealth destruction. Remember: it takes a 100 per cent gain to recover from a 50 per cent loss. As of 2025, investors are once again in a 'bond winter', facing a cumulative real loss of around 30 per cent, driven by the high inflation of the early 2020s and the subsequent rise in bond yields. Stocks: Long-term gain, long-term pain Stocks can really disappoint in both the short term and the long run. Not every dip is followed by a swift recovery. Inflation can further erode real returns. The 21st century alone had three drawdowns of more than 30 per cent in real terms. These huge and frequent losses are a feature of stock markets and thus, most investors are well aware of the short-term risks. Over the long term, equities deliver higher returns than bonds. Yet, over multi-decade horizons, equities can still disappoint. Recent research by Edward McQuarrie suggests that even in the 19th century, stocks did not consistently outperform bonds, challenging the assumption that equities are always the safest long-term investment. Comparing asset classes We examine real losses – the decline in purchasing power – across four key asset classes: savings accounts, government bonds, gold and equities. We look at both short-term (one-year) and long-term (10-year) risk using the conditional value at risk (CVar), a measure of average losses in the worst periods. This measures the expected loss in the worst periods. Savings accounts can quietly erode wealth over time. The accompanying graphic highlights a key paradox – savings are comparatively safe in the short run, but are far from secure over longer horizons. Bonds offered somewhat better long-term performance, but with deeper short-term drawdowns. Gold, often viewed as a safe haven, is volatile in both the short and long run. Despite this, it can still serve as a useful diversifier, particularly when combined with steady or low-volatility stocks. Equities deliver the highest long-term returns, but also the greatest drawdowns. Long-term investors are rewarded – but only if they can endure severe interim declines. These long-term numbers are rarely shown – and for good reason. Most empirical research focuses on short-term, nominal returns, which offer more statistical power but assume investors only care about monthly volatility. When viewed through real, long-term lens, a very different picture emerges. The takeaway is simple. In the long run, all investments are risky. What matters most is not whether you face risk, but how you manage this risk. A middle way via steady stocks Once you have capital, you're exposed to risk – whether you save or invest. The good news is that risk can be reduced through diversification across asset classes: bonds, equities, savings and even gold. This is one of the few 'free lunches' in finance, reducing risk without sacrificing return. Yet, even in a classic 60/40 (60 per cent bonds, 40 per cent stocks) portfolio, most of the risk still comes from equities. There is a better way: reduce stock market risk by focusing on stable companies, sometimes called 'widow-and-orphan' or steady stocks. These firms tend to deliver consistent returns, much like bonds, but with an important advantage – their earnings can grow with inflation. These low-volatility stocks may lag during strong bull markets, but they tend to hold up better during downturns. The second graphic makes a strong case for both diversification and steady stocks. A portfolio fully allocated to steady stocks exhibits similar expected losses as a traditional or classic 60/40 portfolio. Yet, being fully invested in stocks means tail risk which is apparent since the maximum real losses are higher for steady stocks than for the classic 60/40 mix. Therefore, a steady 60/40 portfolio deserves attention. This portfolio invests in defensive equities and has lower downside risk, comparable to the ultra-conservative permanent portfolio (which invests 25 per cent equally in stocks, bonds, savings and gold), but with meaningfully higher returns. The safest move: Lose less Even the safest investment will lose value at some point. But some portfolios lose less, and losing less gives investors the time and confidence to stay invested. Now imagine again being entrusted with $1 million to preserve and grow over the next decade. You now recognise that it's not an easy task but a balancing act. History suggests the best protection comes from diversified investing, including a meaningful allocation to steady stocks. The writer (PhD) is head of conservative equities and chief quant strategist, Robeco. This content has been adapted from an article that first appeared on CFA Institute Enterprising Investor at
